Banking & Finance 2026

Last Updated October 08, 2026

USA

Law and Practice

Authors



A&O Shearman is a global law firm with nearly 4,000 lawyers worldwide. A&O Shearman has one of the largest and most international teams of banking and finance lawyers of any global law firm and one of the leading debt finance practices in the world, known for ground-breaking deals and structures that eventually become market standards. The firm provides clients, including major banks, private credit funds and financial sponsors, with a full-service offering for a broad range of debt finance products for syndicated and private credit transactions, including senior, second-lien and asset-based credit facilities, mezzanine and holdco debt, recurring revenue financings, bridge and bank/bond financings, high-yield bond offerings, securitisation take-outs, debtor-in-possession and exit financings and restructurings. The firm’s global team spans major financial centres, including New York, London, Paris, Amsterdam, Frankfurt, Milan, Madrid, Luxembourg, Singapore, Hong Kong, and Sydney, creating a cross-border network to support its clients’ success.

Following years of heightened leverage levels in the US loan market, and in connection with the passage of the Dodd-Frank Wall Street Reform and Consumer Protection Act post-2008 global financial crisis, US federal regulators issued Interagency Guidance on Leveraged Lending (the “Guidance”) in 2013.

The Guidance imposes certain requirements on regulated lenders and arrangers aimed at promoting sound risk management. Among other things, regulated lenders have to incorporate, as part of their credit risk analysis, a borrower’s ability to deleverage its capital structure during the term of the loan, and to avoid loans that exceed specified leverage levels. Consequently, less heavily regulated non-bank lenders and foreign financial institutions capitalised on this opportunity to increase their market share by providing higher leverage and riskier loans.

After a clear but tempered recovery in 2024 and a turbulent 2025, the US leveraged loan market entered 2026 with resilient headline activity, amidst persistent geopolitical tensions, but with deeper bifurcation by credit quality, sector and use of proceeds. Activity in the first half of 2026 has been dominated by repricings, extensions and refinancings rather than net new M&A/LBO supply, as sponsors remain focused on exits, primarily through IPOs and secondary sales, add-on acquisitions and addressing upcoming maturities. With stubborn inflationary pressures keeping interest rates elevated, funding remains particularly expensive for lower-rated and sector-exposed borrowers, while larger and stronger credits have generally been able to access still-active loan and high-yield markets on more favourable terms.

Geopolitical risk remains elevated in 2026, driven by ongoing conflicts, notably the war involving Iran and continued instability across the Middle East, changing trade policies and persistent tensions between the USA and China. These factors, together with slower-than-expected reductions in benchmark rates and concerns about AI disruption in certain sectors, have contributed to a cautious market tone and a more selective loan market. In the USA, the first-half of 2026 M&A-related loan issuance rose to USD144.8 billion, up 31% from the comparable period of 2025, but sponsor-driven M&A and LBO volume slowed sharply in the second quarter as geopolitical uncertainty and macro headwinds weighed on private equity deal-making, with US private equity deal volume falling 38% quarter-on-quarter and sponsored M&A loan volume declining 54% from Q1 levels.

With respect to US loan documentation, the continued uncertainty in the market has led to an ongoing focus on lender-protective provisions (including provisions intended to prevent future liability management transactions – LMTs), which had been scaled back during the years prior to the last couple of years’ downturns. Global conflicts have also increased focus by lenders on representations and warranties and covenants relating to compliance with sanctions, anti-corruption and anti-money laundering laws.

Companies raising capital continue to access both the loan and high-yield bond markets in 2026, although use of proceeds remain weighted towards refinancing, maturity extension and repricing rather than new-money issuance. Global leveraged loan and high-yield bond issuance reached USD696 billion in the first two quarters of 2026, the busiest pace since 2021, with refinancings accounting for 51% of issuance, while global M&A-driven issuance rose to USD217 billion but remained 37% below the 2021 peak, up 29% from the comparable period of 2025.

In the USA and Europe, high-yield bond issuance totalled USD257 billion through 30 June 2026, a 20% increase over the prior-year pace, as strong investor demand supported opportunistic refinancing and selected acquisition financing. Covenant terms and protections in the leveraged loan market continue to converge with those of the high-yield bond market, while lenders remain highly focused on the quality of collateral and liability-management protections; 92% of the USD Leveraged Loan Index was covenant-lite as of 31 March 2026.

Certain differences remain between leveraged loan and high-yield bond terms. Loans continue to provide weaker “call” protection in connection with voluntary prepayments. Additionally, in capital structures with both leveraged loans and bonds, lenders typically continue to drive the guaranty and collateral structure and control enforcement proceedings given the increased focus on collateral from a loan perspective.

Providers of leveraged loans continue to push to restrict investments in non-guarantor subsidiaries more often than investors of high-yield bonds. Additionally, many loans contain “most favoured nation” (MFN) protections that require an interest rate reset upon the issuance of certain higher-yielding debt, subject to carve-outs that traditionally limit the duration of the MFN and other limitations on MFN as specifically negotiated in the credit documentation.

Finally, some loan provisions are more permissive than those found in bonds, such as:

  • the lack of a fixed-charge coverage governor on the usage of the “available amount” builder basket for restricted payments;
  • allowing amounts in the “available amount” builder basket to build for positive cumulative consolidated net income in a given period without a corresponding deduction for negative amounts in other periods; and
  • permitting the incurrence of debt by “stacking” based on priority (eg, by first incurring junior lien debt in reliance on a secured leverage ratio and then incurring first lien debt in reliance on a first lien leverage ratio), rather than the bond standard secured leverage governor applying to all such secured debt, regardless of priority (at least in the case of unsecured bonds).

In the USA, private credit remains a significant source of corporate lending; the Federal Reserve has described the US private credit market as approximately USD1.4 trillion, similar in size to both the leveraged loan and high-yield bond markets, while still representing about 10% of overall US corporate borrowing; however, private credit’s share of sponsor-backed leveraged buyout financing is significantly larger, having financed the majority of LBOs by number in 2024 and 2025 before ceding market share back to the broadly syndicated market in 2026.

Available data shows that direct lending activity slowed in the second quarter of 2026 as private equity deal-making weakened. Direct lenders continue to provide anchor orders in syndicated deals, buy second-lien or otherwise difficult-to-syndicate tranches, and offer full financing solutions to large companies and private equity sponsors. At the same time, the broadly syndicated market has reclaimed share from private credit in both new LBO financings and in refinancings of existing direct-lending deals.

These lenders are often more flexible, offering higher leverage, committed delayed draw facilities, payment-in-kind interest and capital for parts of the capital structure that are not readily available in the broadly syndicated market, such as preferred equity, holding company (structurally junior) loans or unitranche facilities. They also provide faster execution and greater certainty of terms, as there is no need to obtain ratings, and no marketing process or modification of loan terms during syndication.

In 2026, the dynamic has shifted from a simple expansion story to a more competitive and selective market. Borrowers increasingly compare broadly syndicated and private credit options, direct lenders compete for credits with strong fundamental performance and, increasingly, those with low AI-disruption risk, and private credit providers continue to offer flexible solutions such as delayed draw term loans, unitranche facilities, preferred equity and holdco loans. At the same time, heightened scrutiny of payment-in-kind features, borrower transparency and bank exposures to private credit has increased the emphasis on diligence and documentation quality across the market.

Competition among bank and non-bank lenders continues to drive financing terms, but the 2026 market is more selective than the refinancing-led market of 2024 and the first half of 2025. Larger, higher-rated corporate borrowers have generally been better positioned to obtain pricing concessions, while private equity-backed, lower-rated and software-exposed borrowers have faced a higher risk premium.

Private equity sponsors, as recurring participants in the syndicated and direct loan markets, continue to leverage underwritten borrower-friendly documentation precedents and market-flex protections to preserve execution certainty and ensure new financings are at least as favourable as prior transactions. However, lenders have increasingly pushed back in the current environment, becoming more focused on creditor-friendly protections, particularly where liability management flexibility, restricted payment capacity, unrestricted subsidiary flexibility, investments in non-guarantors or additional debt incurrence could affect recoveries.

Additionally, elevated redemption pressures at some larger private credit vehicles in the second quarter have constrained direct-lending capacity, further reinforcing lender focus on credit quality and collateral protection.

There has also been an increased focus from lenders on provisions aiming to protect lenders against LMTs (as further explained in the following).

Another recent US market trend is the growth of debt financings at the holdco level. Private equity sponsors’ desire to be more competitive in auction processes, and non-bank lenders (and in recent years, bank lenders) that are seeking to deploy additional capital at attractive returns have contributed to such growth.

Holdco financings often include a payment-in-kind interest construct, which enables the opco structure to keep operating without the need to service additional cash interest and amortisation payments. The holdco loans or notes are structurally subordinated to any debt at the opco level and typically do not have recourse to the assets at the opco level. Therefore, the holdco lenders are typically not party to any intercreditor agreement with the opco lenders. Financing with similar features can be achieved through the issuance of preferred equity at the holdco level.

Sustainability-linked lending remains an established feature of the US loan market, although growth has moderated as the product has matured and standards have tightened. Facilities that tie loan pricing to predetermined ESG or sustainability-linked objectives continue to be used where borrowers can identify meaningful, measurable and independently verifiable targets.

Borrowers, under pressure to prove their sustainability and ESG credentials, will often collaborate with a third-party sustainability structuring agent to develop precise ESG benchmarks that will be monitored throughout the loan. Interest rate margins and fees will ratchet up or down depending on performance against pre-set targets. Over time, these benchmarks frequently evolve to become more rigorous. Increasingly, the materiality of the margin ratchets has been criticised by participants who question whether it is enough to motivate change.

In the USA, banks have the option of being chartered by a state government or the federal government under a so-called dual chartering system. Banks, which are chartered by state banking authorities, are primarily subject to the regulations of the relevant state authority and may also be regulated or supervised by the Federal Reserve and/or Federal Deposit Insurance Corporation (FDIC). Banks chartered by the federal government, on the other hand, are subject to regulation by the Office of the Comptroller of the Currency (OCC) and are required to become members of the “Federal Reserve System”. Under federal law, federal and state banks are also required to obtain insurance from the FDIC protecting depositors.

Although alternative credit providers, direct lenders and other non-bank lenders are primarily subject to Securities and Exchange Commission (SEC) rules and regulations, they may also be subject to regulation under the Investment Company Act (ICA) as an “investment company”. However, such lenders are often exempt from many of the ICA’s requirements and regulations.

In 2026, regulatory and supervisory attention to private credit remains elevated as policymakers monitor the migration of corporate lending from banks to non-bank lenders, the growth of bank exposures to private credit vehicles and structural opacity. The immediate systemic risks appear contained, but regulators continue to focus on transparency, valuation, borrower credit quality, use of payment-in-kind features and the interconnections between banks and non-bank financial institutions.

Foreign lenders are subject to (i) the International Banking Act and (ii) the Foreign Bank Supervision Enhancement Act, as well as regulated by the Federal Reserve, whose approval is necessary to establish foreign banking institutions in the USA. Also, foreign banking institutions are required to seek approval from the OCC or state banking supervisor to establish US branches and agencies.

In 2019, the Federal Reserve finalised new regulatory requirements for US subsidiaries of foreign banks. These provided relaxed capital and stress-testing requirements, while also imposing stricter liquidity requirements.

Under US law, regulations for foreign lenders related to granting security interests to, or providing guaranties in favour of, foreign lenders generally do not differ from regulations that apply to domestic lenders.

The USA does not currently impose any foreign currency exchange controls affecting the US loan market, unless a party is in a country that is subject to sanctions enforced by the Office of Foreign Assets Control (OFAC) of the US Department of the Treasury. OFAC administers and enforces economic and trade sanctions based on US foreign policy and national security goals.

As of 2026, this framework remains unchanged; however, heightened geopolitical tensions and the potential for expanded sanctions regimes – particularly in response to ongoing global conflicts and shifting US foreign policy priorities – have led market participants to pay closer attention to OFAC developments. Lenders and borrowers are placing greater emphasis on compliance procedures and due diligence to ensure adherence to evolving sanctions lists, as enforcement activity and regulatory expectations continue to rise.

Loan agreements in the USA traditionally have negative covenants limiting the borrower’s use of loan proceeds to specified purposes as set forth in the loan agreement. Furthermore, US law restricts the use of loan proceeds that are in violation of the margin-lending rules under Regulations T, U and X, which limit financings used to acquire or maintain certain types of publicly traded securities and other “margin” instruments if the loans are also secured by such securities or instruments.

In US syndicated loan financings, an administrative agent is appointed to act on behalf of the lending syndicate to administer the loan. Further, in some secured transactions, a separate and distinct collateral agent is appointed to co-ordinate collateral-related matters. When financings involve numerous series of debt securities or multiple lending groups sharing the same collateral, security interests are sometimes granted to collateral trustees or other “intercreditor” agents to act on behalf of all creditors, with the trust or intercreditor arrangements setting out the relative rights of the various creditor groups.

In the US loan market, lenders can transfer their interest under credit facilities to other market participants through assignments or participations. An assignment is the sale of all or part of a lender’s rights and obligations under a loan agreement, with the assignee replacing the assigning lender for the portion of commitments or loans assigned. As the new “lender of record”, the assignee benefits from all rights and remedies available to lenders thereunder and takes on the obligations of the lenders.

Assignments usually require the consent of the borrower, the administrative agent and, in the case of letter of credit and/or swingline subfacilities, the letter of credit issuers and the swingline banks. Loan agreements commonly provide for some limitations on borrowers’ consent rights during the continuation of any event of default or, increasingly, only during the continuation of a payment or bankruptcy event of default.

Borrower consent is usually not required for assignments to another existing lender (or its affiliate or “approved fund”). Where borrower consent is required, a negative consent mechanism is common: if the borrower does not object within a set period (usually 5–15 business days), consent to assignment is deemed given. In some instances, this deemed consent only applies to assignments in respect of term loans but not revolving facilities. Negative consent and streamlined assignment processes are now common in most large, syndicated facilities, supporting liquidity and efficient secondary trading.

In contrast, participations involve a transfer of limited lender’s rights, which traditionally are focused on the right to receive payments on the loan and the right to direct voting on a limited set of “sacred rights”. The transferee becomes a “participant” in the loan but does not become a lender under the loan documentation and has no contractual privity with the borrower. Participations rarely require notice or borrower consent, but some borrowers have sought to include such requirements though lender resistance remains strong, and most agreements still provide broad flexibility for participations.

There is also a notable trend towards more detailed and expansive “disqualified institution” lists in loan agreements, reflecting borrowers’ heightened focus on controlling the composition of their lender group. These lists generally include the borrower’s competitors, certain undesirable financial institutions (such as vulture and distressed funds) and increasingly, institutions identified for regulatory, reputational or sanctions-related reasons.

Borrowers and their affiliates (including private equity sponsors) are able to purchase loans in the US syndicated loan market, subject to customary requirements and restrictions. This practice remains prevalent, with borrowers and their affiliates continuing to use loan buy-backs as a tool for liability management, opportunistic trading and capital structure optimisation.

In addition, private equity sponsors and their affiliates (other than borrowers and their subsidiaries) are typically allowed to make “open-market” purchases of loans from their portfolio companies on a non-pro-rata basis. These “open-market” purchases are also receiving heightened scrutiny in the context of certain “uptiering” transactions, where “open-market” purchases are used to justify the non-pro rata purchase of loans in the context of exchanging existing debt for new, priming debt. Once held by a borrower affiliate, these loans are normally subject to restrictions on (i) voting, (ii) participating in lender calls and meetings and (iii) receiving information provided solely to lenders.

Loans held by private equity sponsors and their affiliates are also subject to a cap of the aggregate principal amount of the applicable tranche of term loans, traditionally in the range of 25–30%. Bona fide debt fund affiliates of private equity sponsors that invest in loans and similar indebtedness in the ordinary course are usually excluded from these restrictions but are still restricted from constituting more than 49.9% of votes in favour of amendments requiring the consent of the majority of lenders.

There is also increased attention to the definition and scope of “debt fund affiliates” and the mechanics for tracking and enforcing these caps, as the lender base continues to diversify with the growth of private credit, insurance company and asset manager participants. Loan documentation is also evolving to address the treatment of loan purchases by non-traditional investors and to clarify the application of restrictions in the context of complex sponsor structures.

The USA does not have specific rules requiring “certain funds” for acquisition financing. Instead, financing commitments for public and private company acquisitions are typically subject to limited “SunGard” conditions due to the absence of a financing condition in most acquisition agreements, which generally include:

  • accuracy of key “specified representations” about the financing’s enforceability and legality;
  • accuracy of certain material seller or target representations made in the acquisition agreement, breach of which would cause a condition to consummation of the acquisition not to be satisfied or permit the buyer to terminate the acquisition;
  • no material adverse change to the target, matching the condition in the acquisition agreement; and
  • conditions relating to the timing required by arrangers to properly syndicate the loans in advance of acquisition closing (either marketing periods or an “inside date”).

Given these dynamics, it is customary for buyers/borrowers and arrangers to execute commitment letters, including detailed term sheets that usually include the parties agreeing on precedent documentation, simultaneously with signing the acquisition agreement. This provides buyers with committed financing, subject to this customary “limited conditionality”.

Over the past several years, some borrowers facing adverse economic conditions have looked to execute LMTs. The market has also seen a rise in litigation and creditor-on-creditor conflicts arising from aggressive LMT exercises, prompting further evolution in the drafting and negotiation of credit agreements.

One example of such a transaction involves a borrower seeking the release of guarantors that are no longer wholly owned by the borrower (even if wholly owned by its other affiliates). Following such release, the released entities would more easily be able to incur additional indebtedness. Lenders have increasingly sought protection from this type of transaction by permitting the release of a guaranty only in certain circumstances (eg, the guarantor becomes non-wholly owned in a bona fide transaction involving a third party without the intent of releasing the guaranty as part of the transaction).

Another example is the use of multiple-step processes (where each step is permitted under the investment and disposition covenants) to move valuable IP and other assets from guarantors to non-guarantor entities, thereby automatically releasing the lenders’ security interest in such assets in the process. More recently, borrowers have expanded this playbook beyond subsidiary structures altogether. In February 2026, Xerox completed a USD450 million “non-subsidiary drop-down”, contributing its trade marks and other IP to a newly formed joint venture in which it holds no majority voting control, thereby sidestepping the LMT blockers that govern transfers only to unrestricted or non-guarantor restricted subsidiaries. In response, lenders have sought to limit or eliminate this flexibility. Notably, in 2025, the Fifth Circuit held in the Serta Simmons litigation that the 2020 uptier transaction was not a valid “open-market purchase” under the credit agreement, reinforcing lender protections against non-pro-rata priming transactions.

Borrowers have also increasingly used the amendment section to make updates to credit documentation to benefit majority lenders over minority lenders, such as allowing majority lenders subordinate existing debt in both right of payment and on the liens for new debt that they provide. In response, lenders are now tightening amendment provisions to require any priming debt to be offered to all lenders pro rata. Lenders are focused on the “pro rata sharing” provisions as well as enhanced protections for minority lenders.

Some borrowers have also raised new secured debt outside the existing credit group, which is on-lent into the credit group on a secured basis, and is secured and guaranteed by the existing credit group (on a pari passu basis with the existing debt), giving new lenders two secured claims against the existing credit group and increasing their pro rata share of potential recovery proceeds vis-à-vis the existing creditors. Lenders are pushing for stricter limits or bans on unrestricted subsidiaries holding debt or liens on any property of the borrower and the restricted group to prevent dilution of recovery in a distressed scenario.

Additionally, lenders are concerned about value leakage from the credit group through investments in unrestricted subsidiaries. Borrowers have moved valuable assets outside the reach of existing lenders to incur new, structurally senior debt secured by those assets, thereby undermining the collateral and credit support available to the original lender group. To address this, lenders are demanding tighter controls on investments in unrestricted subsidiaries, such as requiring such investments be made solely through specific baskets, prohibitions on the reallocation or reclassification of investment capacity to unrestricted subsidiaries and aggregate caps on the value of assets that can be transferred.

Although not a US regulation, CRD VI is increasingly relevant to US banking groups that arrange cross-border lending, guarantees or other core banking services into the EU. Article 21c of CRD VI (Directive (EU) 2024/1619) establishes an EU-wide third-country branch perimeter for core banking services, including deposit-taking, lending and guarantees, generally requiring a non-EU bank to operate through an authorised branch in the relevant member state or an authorised EU subsidiary unless an exemption applies. Key exemptions and carve-outs include reverse solicitation, intragroup arrangements, certain interbank activity and Markets in Financial Instruments Directive (MiFID)-related ancillary services. US banks have not converged on a collective approach amid the absence of clear regulatory guidance. The Loan Syndications and Trading Association (LSTA) has, however, issued a market advisory on Capital Requirements Directive (CRD) VI (10 July 2026) providing initial guidance for loan market participants.

Nationally chartered banks may not charge interest exceeding the greater of (i) the rate permitted by the state in which the bank is located or (ii) 1% above the discount rate on 90-day commercial paper in effect in the bank’s Federal Reserve district.

If the state where the bank is located does not prohibit usurious interest, banks may not charge interest exceeding the greater of 7% or 1% above the discount rate on 90-day commercial paper in effect in the bank’s Federal Reserve district. In general, federal law will pre-empt any state usury law that prohibits state-chartered banks from applying the same interest rate as a nationally chartered bank.

Under New York law, with certain exceptions, charging interest in excess of 16% constitutes civil usury, and charging interest in excess of 25% constitutes criminal usury. However, loans in excess of USD250,000 are exempt from the civil statute, but remain subject to the criminal statute. Loans in excess of USD2.5 million, which include nearly all broadly syndicated loans in the USA, are exempt from both New York’s civil and criminal statutes.

There are no rules or laws in the USA that prohibit certain disclosure of financial contracts, but in credit documentation there is traditionally a confidentiality section that prohibits the lenders from disclosing the nature of the financing other than in pre-agreed situations.

The US tax rules contain a complex withholding regime that imposes, in certain circumstances, a withholding tax of up to 30% on payments of interest to non-US lenders. To encourage international lending to US borrowers, however, the rules contain various exemptions from this withholding tax. Under current law, the expectation is that lenders to a US obligor should generally be able to qualify for one or more of these exceptions, such that lenders are not subject to the withholding tax and obligors are not required to compensate lenders under a “gross up” provision in credit agreements. To benefit from these exemptions, however, lenders must provide certifications to borrowers or their agents, generally on tax forms published by the Internal Revenue Service (IRS), as discussed in the following. Parties to credit agreements with US obligors should ensure that such forms are appropriately addressed in loan documentation and furnished in practice.

This withholding tax regime may also apply to certain other payments and income arising from loans. If a loan is issued at a discount in excess of a de minimis amount (original issue discount – OID), this discount is treated as interest income when paid, subject to the withholding tax. Certain fees may also be treated as OID for this purpose.

There are several exemptions from the withholding tax on interest. The most notable exemption availed by non-bank lenders is the portfolio interest exemption. In the case of banks and other lenders that do not qualify for the portfolio interest exemption, US tax treaties may eliminate withholding or reduce the rate. Finally, if non-US banks lend from their branch in the United States (a “US trade or business”), the withholding tax generally does not apply.

To qualify for one of these exemptions, non-US lenders are generally required to provide a US tax form to the borrower or agent – usually an IRS Form W-8BEN-E (for treaty benefits or the portfolio interest exemption) or IRS Form W-8ECI (if the interest is effectively connected with the non-US lender’s US trade or business). Additional certifications and forms are required in certain instances involving flow-through entities or intermediaries.

Another withholding regime that may apply to certain payments of interest and OID is the “backup withholding” regime, which generally applies to domestic payments (currently at a withholding rate of 24%) in circumstances where a US lender fails to provide certain information and certifications required for purposes of the US information reporting regime. Backup withholding is usually eliminated by the provision of an IRS Form W-9 and, if it is imposed, generally can be recovered in the form of a credit on the lender’s US tax return.

Principal payments and proceeds from a sale or other disposition of debt instruments are not subject to US withholding tax (except to the extent that such payments are treated as a payment of interest or OID). However, fee income that is not treated as OID may be subject to 30% withholding unless a treaty applies or the recipient is engaged in a US trade or business. The portfolio interest exemption may not apply to such fees because they may not be treated as interest for US tax purposes.

Finally, the Foreign Account Tax Compliance Act (FATCA) may impose a 30% US withholding tax on non-US banks and financial institutions (including hedge funds) that fail to comply with certain due diligence, reporting and withholding requirements. FATCA withholding tax applies to payments of US-source interest and fees, without any exemptions for portfolio interest or treaty benefits. Following the guidance issued by the IRS and US Department of the Treasury in 2018, FATCA no longer applies to payments of gross proceeds from a sale or other disposition of debt instruments of US obligors. In the case of payments that are within FATCA’s purview, the recipient must generally certify its compliance with FATCA in order to avoid a punitive 30% withholding tax (on the same IRS W-8 forms described previously).

Many countries have entered into agreements with the USA to implement FATCA (intergovernmental agreements – IGAs), which may result in modified requirements that apply to financial institutions organised in such countries.

Under Section 956 of the Internal Revenue Code, if a foreign subsidiary of a US borrower that is a controlled foreign corporation (CFC) guarantees the debt of a US-related party (or if certain other types of credit support are provided, such as a pledge of the CFC’s assets or a pledge of more than two-thirds of the CFC’s voting stock), the CFC’s US shareholders could be subject to immediate US tax on a deemed dividend from the CFC.

Following regulatory changes published by the US Treasury and the IRS in 2019, US borrowers may obtain credit support from CFCs without incurring additional tax liability if certain conditions are met. However, despite these regulatory changes, the majority of loan documents today continue to maintain customary Section 956 carve-outs. This excludes CFCs from the guaranty requirements and limits pledges of first-tier subsidiary CFC equity interests to less than 65%.

Separately, non-US lenders should closely monitor their activities within the USA to determine whether such activities give rise to a US trade or business or a permanent establishment within the USA. If so, they could be subject to US taxation on a net income basis.

The primary tax concerns that arise for non-US lenders to US obligors are those summarised in 4.1 Withholding Tax; ie, withholding tax on interest, including FATCA withholding. To mitigate these concerns, it is important for non-US lenders and US obligors to ensure that appropriate tax forms are exchanged in order to establish any exemptions from these withholding regimes.

Although the US tax rules do not address non-money centre banks per se, the various regimes described in 4.1 Withholding Tax (including IRS tax forms, the portfolio interest exemption and FATCA) apply differently and impose different requirements based on the particular circumstances and business activities of the lender.

In the United States, secured financings typically require a collateral package consisting of substantially all assets of the borrowers and their subsidiaries, along with a pledge of the borrower’s equity interests by its direct parent – with negotiated exceptions, which typically excludes assets with burdensome perfection requirements and/or where a pledge would lead to expensive or other negative consequences for the borrowers that outweigh the benefit to the lenders. Typical exclusions include:

  • leased and owned real property below an agreed threshold;
  • equity interests in certain non-guarantor subsidiaries;
  • contractual rights that cannot be pledged by law or contract (although proceeds thereof are generally included);
  • assets requiring third-party or government consent to pledge;
  • assets of de minimis value;
  • assets that would cause negative tax consequences if pledged;
  • assets subject to burdensome perfection requirements, such as certificates of title; and
  • “intent-to-use” trade mark applications for the registration of a trade mark.

The creation of security interests for most categories of personal property is governed by the Uniform Commercial Code (UCC). The requirements for creating enforceable security interests with respect to personal property under Article 9 of the UCC are as follows:

  • the lender must provide value to the grantor of the security interest;
  • the grantor must have rights in the collateral or the power to transfer rights in the collateral to the lender; and
  • either the grantor must execute a security agreement, which must be authenticated by the grantor and describe the collateral, or, in the case of certain types of collateral, the collateral must be in the possession or control of the lender.

To create a security interest in assets not governed by the UCC (eg, real property and certain kinds of IP), the parties will typically create separate collateral documents or mortgages pursuant to applicable legal requirements in the jurisdiction governing the property.

Lenders must perfect such security interest to obtain priority vis-à-vis other creditors. The relevant perfection requirements under Article 9 of the UCC depend on the asset type, but generally Article 9 of the UCC provides the following four methods of perfecting security interests in domestic personal property.

  • Filing a UCC-1 financing statement in the appropriate jurisdiction (which is a short document setting forth basic information about the grantor and the secured party, and a collateral description).
  • Possession, in the case of certain tangible assets (such as “certificated securities”).
  • Establishing control, which may be effected by entering into control agreements in the case of deposit accounts, letter of credit rights, investment accounts and electronic chattel paper. In addition, the 2022 amendments to the UCC introduced Article 12, which governs “controllable electronic records” (including certain digital assets and electronic money) and provides a new perfection method through “control” for these emerging asset categories; a growing number of states have adopted these amendments.
  • Perfection upon attachment (ie, automatically upon the creation of the security interest), for certain other personal property.

Perfection of security interests in federally registered copyrights (and, by custom, patents and trade marks) requires filing with the US Copyright Office (or the US Patent and Trademark Office), in accordance with federal law. Various state and federal laws govern the perfection of security interests in motor vehicles, aircraft, ships and railcars, with separate registries and perfection steps. Mortgages in real property are perfected by recording such mortgages (or equivalent documents) with the local (usually county-level) recording office where the real property is located.

Article 9 of the UCC permits the granting of a floating lien in the form of an “all assets” pledge, which can include all personal property owned by the grantor. Further, there is no distinction between floating and fixed charges in the USA, so the granting of security interests over personal property normally covers both presently owned and later acquired assets. Importantly, however, “all assets” pledges apply only to personal property that is subject to the requirements of Article 9 of the UCC (with certain exceptions for asset types such as commercial tort claims, which must be described with more specificity).

Other assets – such as real property and federally registered copyrights – cannot be subject to floating liens. For certain asset types, such as motor vehicles, creation of a security interest is governed by Article 9 of the UCC, but perfection is governed by state certificate of title laws, so perfection of security interests over such assets cannot be obtained by filing a UCC-1 financing statement.

In the USA, there are generally no limitations or restrictions on the provision of guaranties to related parties. However, to avoid a guaranty from being rendered unenforceable on the grounds of fraudulent conveyance, downstream, upstream and cross-stream guaranties should include a limit on the guaranteed amount and the guarantor must either receive adequate consideration or must not be rendered insolvent after giving effect to such guaranty. Customary limits contained in guaranties are designed to prevent the guarantor from being rendered insolvent. In addition, loan market participants often require borrowers and their subsidiaries to provide certifications as to their solvency at the time the loan and the guaranties thereof are made.

There are no US rules generally prohibiting a target company from guaranteeing or granting a security interest in its assets to provide credit support for a financing used to acquire its or any of its parent entities’ shares. However, guaranties and security interests provided by a target company are subject to the rules on fraudulent conveyance and, in certain cases, may be subject to regulatory schemes that make such a guaranty and/or security interest impracticable even if legal.

Subject to such limitations, lenders will typically require guaranties and security interests to be provided by the target company – along with delivery of any certificated securities of the target company – as a condition to the closing of an acquisition financing subject to any limits that “SunGard” provisions impose.

Anti-assignment provisions in commercial contracts pose difficult issues for lenders in secured financings. A statutory override of anti-assignment provisions in contracts is generally available under the UCC but, if the restricted collateral is critical to the collateral package, lenders are likely to require such third party to consent to the pledge as a condition to the loan to prevent future enforcement uncertainties.

US loan documentation typically allows releases of the lenders’ security interest in collateral in connection with dispositions of such collateral that are permitted under the loan documentation. The release of all or substantially all of the collateral typically requires the consent of all lenders or, in some cases, a super majority thereof.

The UCC of the applicable jurisdiction governs the relative priority of security interests held by different creditors in the same assets of a grantor and is subject to the following rules:

  • a perfected security interest has priority over a conflicting unperfected security interest;
  • conflicting perfected security interests rank in priority according to the time of filing or perfection; and
  • conflicting unperfected security interests rank in priority according to the time at which the security interest attached or became effective.

In addition, the UCC allows certain categories of collateral to be perfected by multiple methods, with priority determined based on the “preferred” method, regardless of the aforementioned rules. With respect to investment property, securities accounts and certificated securities, perfection via “control” or possession has priority over perfection via filing a UCC-1 financing statement. Further, the UCC contains an exception for purchase money security interests under which a secured creditor with a purchase money security interest can obtain priority ahead of an earlier UCC-1 financing statement with respect to the purchased asset(s).

Lenders and borrowers are allowed to agree to modify the priority rules set out in the UCC and other relevant laws, by contract. The parties can also accomplish different lien priorities structurally.

Arrangements for lien subordination ordinarily provide that:

  • junior creditors are subject to a “standstill” period prior to exercising enforcement rights or remedies with respect to shared collateral;
  • payments from the proceeds of shared collateral received by junior creditors in violation of the agreement will be held in trust and turned over to senior creditors; and
  • certain specified amendments to both senior and junior priority loan documents will be subject to agreed limitations.

Structural subordination arises where obligations incurred or guaranteed solely by a borrower are effectively junior to obligations incurred or guaranteed by a subsidiary of the borrower, to the extent of that subsidiary’s assets. In any insolvency scenario, the subsidiary’s creditors have the right to be repaid by such subsidiary (or out of its assets) as direct obligations of such entity before creditors of the parent borrower – such subsidiary’s equity holder – are repaid. Where the parent borrower is primarily a “holding company” for the equity interests of its operating subsidiaries, creditors of an operating subsidiary will be paid in full in priority to the holding company’s creditors from assets of such subsidiary.

Mechanic’s liens arise when contractors, subcontractors or suppliers are unpaid for work performed or materials supplied. This lien is a security interest in the property. If the owner tries to sell the property, the lienholder will have a secured interest in the portion of the proceeds needed to pay the debt.

Tax liens are placed against property by the local, state or federal government, as authorised by statute, for delinquent taxes, including property, income and estate taxes. A judgment lien is any lien placed on the defendant’s assets as a result of a court judgment.

Possible structuring concerns will focus on properly conducting due diligence on any possible liens, including conducting searches and other disclosure requirements as set forth in the credit documentation.

Loan and security documentation generally provides a customary set of enforcement rights and remedies to secured parties, exercisable by such parties following the occurrence of a “default event” by an obligor.

From a statutory perspective, Article 9 of the UCC gives secured parties the right to proceed with several enforcement methods after a default event has been triggered by an obligor. These rights include:

  • the right of a secured party to collect payments directly from a third-party obligor under accounts receivable, deposit accounts or with respect to certain other types of intangible assets;
  • the right of a secured party to repossess collateral, either through the institution of judicial proceedings or through a non-judicial action; and
  • the right of a secured party to dispose of the collateral through a public or private sale process.

However, to exercise such remedies under Article 9, secured parties also have an obligation to comply with certain statutory requirements. Such requirements are designed to protect obligors and generally provide that:

  • the time, place and/or manner of exercising such remedy must be commercially reasonable;
  • sufficient advance notice is provided to the relevant obligor; and
  • certain other creditors who have an interest in the collateral are given adequate notice where such sale process involves a public sale or auction.

New York courts generally permit parties to a loan agreement to select a particular foreign law to govern their contract unless the choice of law conflicts with public policy or there is no reasonable basis for the parties to choose such law to govern their contract (ie, the law selected has no real relationship to the parties or the transaction); then the courts may decline to enforce the selected governing law.

In terms of conflict of laws rules, New York’s rules will generally uphold foreign forum selection clauses so long as the selected jurisdiction has a reasonable relationship to the transaction – more specifically, a significant portion of the agreement was negotiated, or the agreement was substantially performed, in such jurisdiction.

In cases involving foreign states, the Foreign Sovereign Immunities Act will permit a waiver of immunity either explicitly or by implication.

Subject to certain conditions being observed (including due process requirements and reciprocity), New York courts will generally recognise and enforce the judgments of foreign courts. However, although uniform laws have been adopted by many US states, when recognition and enforcement of foreign judgments are concerned, there is still significant diversity between the states when dealing with procedural and substantive considerations.

A foreign lender’s ability to enforce its rights under a loan or security agreement will depend on the facts and circumstances of each case.

Automatically upon the filing of a petition to commence insolvency proceedings under the United States Bankruptcy Code, an “automatic” stay comes into effect, prohibiting perfection of interests, termination of contracts and enforcement activities by creditors, with few exceptions. This stay prevents the proverbial creditor “race to the courthouse” and provides the debtor with a “breathing spell”, typically to organise a sale or a plan of reorganisation or liquidation.

Lenders’ enforcement rights are replaced with rights in the bankruptcy case, and lenders may seek repayment from sale proceeds or estate distributions, which may take a variety of forms, including payment of cash or equity, reinstatement of debt and issuance of replacement obligations. In chapter 11, the reorganisation chapter of the Bankruptcy Code, individual creditors are entitled to recover the liquidation value of their claims regardless of how similar creditors placed in the same class vote.

Classes of creditors may be bound to a plan when ⅔ in amount and more than 50% in number of the class approve. Class approval is not required; however, the “cram down” standards must be met. These generally prohibit distributions to junior creditors or equity where senior dissenting classes are impaired and require that secured creditors either receive their collateral, its proceeds or its "indubitable equivalent” value, or secured replacement notes. Chapter 7, the liquidation chapter of the Bankruptcy Code, has its own distribution rules.

Secured creditors also have the right to credit bid in a sale of their collateral, must consent to the use of cash collateral unless their security interest is “adequately protected” and may seek adequate protection against diminution of the value of their collateral, or relief from the automatic stay for cause.

The Bankruptcy Code recognises certain rights of lien and payment priority, which are set out in broad strokes as follows.

  • First, secured creditors are paid from the value of their collateral, subject to estate claims for the costs of maintaining such collateral.
  • Then come administrative claims, priority claims, general unsecured claims (including deficiency claims of undersecured creditors) and equity, in that order.
  • Administrative claims include expenses of administering the estate, operating the business on a post-petition basis and certain statutorily designated items (eg, claims for goods delivered within the 20 days before the petition date and claims arising from a failure of adequate protection).
  • “Priority” general unsecured claims include, among other things, certain taxes and employee claims. Administrative and priority claims must be paid under a plan (some priority claims can be paid over time), and certain priorities apply within these categories.

Case length depends on a variety of factors, with the most important being the level of advance planning and creditor agreement at the petition date. The trend in recent years has been for debtors and their stakeholders to undertake greater planning in advance of filing for bankruptcy and seek to achieve specific objectives while in chapter 11, which results in shorter cases.

In chapter 11 cases, if creditors are solicited on a “prepackaged” plan and certain notice periods are permitted to run prior to filing the case, a bankruptcy case can be as short as a day. More typically, prepackaged bankruptcies take 45–60 days from the petition date.

“Prearranged” plans, where requisite creditors have largely agreed to a plan framework but have not been solicited before the case is filed, can also be expedited and completed within two to three months.

If a plan must be formulated after the filing, three to six months is more typical, and cases with more complex issues, litigation, mass torts liabilities or lack of consensual resolution can take significantly longer. There is no time limit for exiting bankruptcy, but the debtor may not maintain the exclusive right to file a plan for longer than 18 months after the petition date and/or the exclusive right to solicit a plan for more than 20 months after the petition date. Cases may also be dismissed or converted to liquidation, particularly where there is no prospect of reorganisation.

Chapter 11 effectively preserves going-concern value, and sizeable enterprises frequently reorganise successfully using this process. Additionally, there is a mature investor base specialised in acquiring distressed companies (or their debt or assets), which aids in supporting value. Companies that cannot reorganise may be liquidated under Chapter 7.

Borrowers and significant creditors have increasingly sought ways to address liquidity concerns or restructure without commencing bankruptcy proceedings. These approaches are often preferred by secured creditors because they avoid the significant costs associated with bankruptcy proceedings and provide a more streamlined process to obtain control of the borrower.

Where the borrower can obtain requisite consents, parties may restructure debt or other obligations without bankruptcy proceedings; for example, borrowers and issuers may seek to:

  • exchange or amend existing debt to allow for covenant relief;
  • extend payment terms or payment relief; and
  • sometimes simultaneously solicit consents for a pre-packaged bankruptcy to be filed if requisite consents are not obtained.

Inducements can be offered as well, such as improvements in collateral, guaranties, or other terms. As described in more detail in 3.7 Debt Buybackand 3.9 Recent Legal and Commercial Developments, borrowers are increasingly utilising LMTs to obtain more liquidity and flexibility in order to avoid or delay a deeper restructuring or chapter 11. Some deals involve incumbent lenders providing additional capital or other concessions to participating lenders in exchange for improvements in their priority position relative to other lenders or by lending against separate collateral. The ability to effectuate such transactions without unanimous consent allows the architects of these transactions to propose coercive terms that leave non-participating lenders in a worse collateral, guaranty and/or covenant position, or to exclude some lenders from participating altogether.

Secured lenders are increasingly resorting to the use of out-of-court solutions when faced with defaulting or distressed borrowers. These can take the form of out-of-court M&A processes run in co-ordination with a borrower or the exercise of more formal remedies, with or without borrower co-operation. When there are few secured creditors in the capital structure and no need to restructure operations, consensual or non-judicial foreclosures under Article 9 foreclosures of the UCC are relatively common. Some equity investors can also handle operational restructurings without the tools of bankruptcy.

Recently, some debtors facing the need to restructure have sought to do so through foreign insolvency proceedings rather than through chapter 11 reorganisations. The last year has seen a number of US-based companies use UK restructuring plans, which can be less disruptive to a company’s operations, require a lower level of consent from creditors and can be completed more quickly and at lower cost. Under certain circumstances, US bankruptcy courts will recognise foreign insolvency proceedings and enforce a foreign restructuring plan, even if it provides for remedies or transactions that are not available in chapter 11.

Insolvency of an obligor creates risks of a change of control, degradation of the value of the obligors or their assets and avoidance liability. In a bankruptcy, the company may be sold or transferred to creditors regardless of any constraints in loan documentation. In some scenarios, lenders, including secured lenders, may be given notes against the reorganised company. Additionally, lenders can be forced to accept virtually any distributional outcome that provides more than liquidation value if their class consents.

Any circumstance that further stresses the business, creates a forced-sale dynamic or delays the process can diminish recoveries. Dilution by other creditors, including priming financing and related fees, additional equity financing provided under a rights offering and related fees (often in the form of rights to acquire equity at a discount), distributions to senior creditors under a low valuation and necessary payments to other creditors, as well as the costs of the process, are all potential causes of lost value.

Finally, depending on the timing and circumstances of their loan, some lenders may be subject to risks of avoidance of rights transferred to them or obligations undertaken by the estate. The most typical of these is preference liability for transfers to unsecured or under-secured creditors within the 90 days preceding the case on account of antecedent debt. Fraudulent transfer liability generally arises in circumstances where a debtor is insolvent or inadequately capitalised and does not receive reasonably equivalent value for a transfer or obligation, or where the transfer is intended to hinder creditors.

Project finance remains widely used in the USA, especially in renewable energy, digital infrastructure, mining and public-private partnerships (PPPs). The growth of AI-related data centre development has increased demand for large-scale power, real estate, fibre and infrastructure financing, while renewable energy and mining projects continue to rely on a mix of project finance and corporate finance techniques.

There is also growing use of hybrid project finance-corporate finance structures for portfolios of projects, where multiple assets are grouped and secured together. In mining, alternative sources of funding such as streaming, royalty and/or prepay contracts are now common, but a portion of the project funding typically continues to be provided under a traditional project finance structure.

More US projects have been successfully procured as P3s, relying on a “user fee” or an “availability payment” model and utilised in transportation, social infrastructure, water/wastewater, energy (particularly at universities) and telecom/broadband sectors. While availability payment structures are most common, the influx of infrastructure funds and private equity is driving value creation through risk or commercialisation.

The Infrastructure Investment and Jobs Act is expected to support P3s by authorising USD550 billion of new federal investments in infrastructure projects, renewing the Transportation Infrastructure Finance and Innovation Act (TIFIA), Railroad Rehabilitation and Improvement Financing (RRIF) and Water Infrastructure Finance and Innovation Act (WIFIA) loan programmes, doubling the cap on private activity bond (PAB) issuance to surface transportation projects and directing the Secretary of Transportation to establish a programme to enhance public entities’ technical capacity to facilitate/evaluate P3s.

Project documents in US financings are often, but not required to be, governed by local law. Construction law is state-specific; therefore, best practice often leads to the signing of construction contracts under local law. Project documents use a mix of submission to jurisdiction to local courts and arbitration to resolve disputes based on the characteristics of the project, bargaining power of the parties and other factors.

Various state and federal laws limit or prohibit non-resident foreign persons or entities controlled by such persons from acquiring US real property or impose reporting requirements on foreign owners of US real estate. Many of these laws apply only to mineral resources or to agricultural property. Before acquiring real estate, a purchaser is advised to review the relevant federal laws that apply to the prospective purchaser and the type of property being acquired, and to consult with counsel in the state in which the property is located for an understanding of the relevant state laws.

The main issue is whether the transaction will be a limited recourse deal or whether there will be a completion or other form of sponsor guaranty. This is particularly relevant in energy transition deals, such as hydrogen deals, where limited recourse financing may not be readily available. Another key issue when structuring the deal is to consider whether tax equity will be used to finance the project, as this will impact the terms of the project finance debt.

The federal and state laws relevant to project companies vary depending on the sector. In the case of the energy industry, key federal statutes include the Federal Power Act, the Public Utility Regulatory Policies Act of 1978 (PURPA) and the Public Utility Holding Company Act of 2005 (PUHCA).

In their simplest form, project financings are provided by syndicates of commercial banks, often together with development financial institutions (DFIs) and export credit agencies (ECAs) for projects in emerging markets. These financings are structured as senior secured financings with a first lien on all project assets/equity with limited or no recourse to the project sponsors.

However, project financings are becoming increasingly complex multisource financings in which commercial banks and DFIs/ECA facilities combine with private equity, commodity traders, strategic investors (original equipment manufacturers – OEMs), governmental entities, project bonds, ESG and streaming/royalty company funding in the form of debt, prepaid forwards, leases, concessionary facilities, grants and hybrid debt/equity facilities, to name a few available investment instruments.

A key issue with US natural resource projects, particularly mining projects, is the lengthy and challenging permitting process. A key consideration associated with downstream projects in the sector is the availability for the particular project of benefits afforded by the Inflation Reduction Act, whether in the form of tax credits, grants or concessionary loans. The availability of such benefits can significantly enhance the financial feasibility of a project.

The principal environmental laws include:

  • the US Comprehensive Environmental Response, Compensation and Liability Act, which governs the clean-up of soil and groundwater contamination and includes a “lender liability exemption”;
  • the Resource Conservation and Recovery Act, which requires “cradle-to-grave” management and disposal of waste;
  • the Clean Air Act;
  • the Clean Water Act; and
  • the Emergency Planning and Community Right-to-Know Act, which requires industry to report on the storage, use and release of certain chemicals to federal, state and local governments (overseen by the US Environmental Protection Agency).

The principal health and safety law is the US Occupational Safety and Health Act, which is overseen by the Occupational Safety and Health Administration.

With the new US administration, there is heightened uncertainty regarding the direction of federal environmental policy. Market participants should monitor for changes in EPA enforcement priorities, as the administration may seek to adjust the balance between environmental regulation and economic growth, particularly in sectors such as energy, infrastructure and manufacturing.

*The authors would like to give a special thanks to Maura O’Sullivan, debt finance partner at A&O Shearman, for her contributions to this article.

A&O Shearman

599 Lexington Avenue
New York, NY
10022-6069
United States of America

+1 212 848 4000

mosullivan@aoshearman.com www.aoshearman.com
Author Business Card

Trends and Developments


Authors



DLA Piper is a leading global law firm with more than 5,500 lawyers across over 90 offices in more than 40 countries, providing clients with seamless legal services across jurisdictions and industries. The firm advises multinational corporations, financial institutions, private equity sponsors, investment funds, and emerging businesses on their most significant transactions, disputes, regulatory matters, and strategic initiatives. Recognised for its deep local market knowledge and integrated global platform, DLA Piper delivers co-ordinated, cross-border solutions designed to help clients navigate growth, transformation, and complex commercial challenges. The firm is committed to innovation in legal service delivery, leveraging technology, collaboration, and efficient staffing models to provide consistent value and exceptional client service. DLA Piper also maintains a strong commitment to pro bono work, sustainability, diversity and inclusion, and community engagement. Through its collaborative approach and global reach, the firm serves as a trusted advisor to clients operating in an increasingly interconnected business environment.

How Artificial Intelligence Is Transforming Fund Finance: A Lender’s Perspective

Introduction: two speeds of adoption

Artificial intelligence (AI) is reshaping private credit and fund finance, but adoption is happening at two markedly different speeds, and the gap between them presents both the opportunity and the challenge facing the market today.

Private credit firms have moved rapidly, embedding AI into deal screening, diligence, benchmarking and portfolio reporting as a workflow-native tool. For these firms, AI is not an innovation project or a technology experiment – it is already part of daily operations. Industry surveys show substantial planned investment increases in data automation, and a large majority of institutional investors are already using or planning to use generative AI for private markets.

Understanding the structural factors driving this adoption gap may be useful for fund finance practitioners navigating the current landscape. Private credit firms operate with significant regulatory flexibility, allowing them to experiment with AI tools without the same supervisory scrutiny that constrains banks. Their technology investments face shorter approval cycles and less intensive governance review. This agility advantage may compound over time as early adopters accumulate experience, refine their workflows and build institutional expertise.

Banks present a fundamentally different picture shaped by their regulatory obligations. They have deployed traditional AI for years in fraud detection, anti-money-laundering compliance, credit scoring and transaction monitoring, but they adopt generative AI more deliberately.

The reasons are structural rather than cultural: model risk requirements, explainability obligations, data privacy constraints and supervisory expectations that do not apply to unregulated lenders. The vast majority of large banks already use some form of AI, yet the gap between traditional and generative deployment remains substantial and will likely take time to close.

This two-speed dynamic creates both opportunity and challenges in the fund finance sector. Fund finance products – subscription credit facilities, net asset value facilities and hybrid structures – are intensely document-driven. They involve repetitive extraction of terms from limited partnership agreements (LPAs), side letters and investor documentation – precisely the tasks where AI delivers significant efficiency gains when properly supervised. Governance maturity, not raw speed, will ultimately distinguish market leaders from institutions left behind by the pace of change. The question for most institutions is not whether to adopt AI, but how to do so at a pace that matches their governance capacity and risk appetite.

The competitive implications of this adoption disparity may extend beyond individual deal economics to market positioning. Institutions that effectively integrate AI into their fund finance operations may be able to process larger deal pipelines with existing headcount, respond more quickly to financing requests and maintain more consistent documentation standards across their portfolios. The efficiency differential between AI-enabled and traditional workflows appears to be most pronounced in documentation-heavy transactions where extraction, comparison and verification tasks dominate the professional time allocation.

The regulatory landscape: institutional guardrails, not statutory mandates

The regulatory environment for AI in banking remains transitional and, for many institutions, frustratingly unclear in its practical implications. In the United States, banking regulators have historically applied technology-neutral supervisory expectations to AI systems, treating them within existing model risk management frameworks without creating AI-specific rules. This approach reflected a reasonable assumption that AI was simply another form of quantitative modelling – subject to validation, testing and governance like any other analytical tool deployed within a financial institution.

That assumption is now being actively revisited. In April 2026, the Federal Reserve Board, the Office of the Comptroller of the Currency and the Federal Deposit Insurance Corporation issued revised interagency guidance on model risk management (SR 26-2; OCC Bulletin 2026-13), superseding the 2011 guidance that had governed model risk management practices for 15 years. The revised guidance narrows the definition of a “model” and expressly places generative and agentic AI outside its scope on the basis that those technologies are “novel and rapidly evolving”.

The carve-out is not an exemption. The agencies stated that a banking organisation’s general risk management and governance practices should determine the controls applied to tools that fall outside the guidance, and they have announced a forthcoming request for information addressing banks’ use of AI, including generative and agentic AI. The practical consequence is a governance gap rather than a governance holiday: systems that materially influence credit and portfolio decisions may sit outside formal model validation while remaining squarely within safety and soundness expectations.

No comprehensive federal AI legislation for financial institutions has been enacted as of mid-2026, though AI-specific guidance has been signalled as forthcoming. Senior Federal Reserve officials have publicly acknowledged that generative AI “may require a different approach” from existing model risk frameworks. In the European Union, certain aspects of the AI Act’s phased implementation timeline have been extended, and the European Central Bank maintains a technology-neutral supervisory posture – focusing on how institutions apply and govern AI rather than prescribing specific technology requirements or mandating particular architectures.

The trajectory of regulatory development suggests increasing specificity in supervisory expectations, even absent formal rulemaking. Examination teams are asking more detailed questions about AI governance, and consent orders in unrelated enforcement matters increasingly reference technology oversight as a supervisory priority. Institutions that develop comprehensive AI governance frameworks now position themselves to demonstrate compliance if more prescriptive guidance emerges, potentially reducing the need for later modifications to governance processes.

The practical implication for fund finance lenders is significant and worth stating directly: the guardrails today are institutional, not statutory. A bank’s own governance framework – its model validation protocols, data handling policies, human oversight requirements, escalation procedures and documentation standards – currently serves as the design specification for AI deployment. Institutions with mature governance frameworks can move more confidently into AI-assisted workflows; those without them face self-imposed constraints that may prove more restrictive than any regulation that eventually emerges from the ongoing supervisory deliberations.

For fund finance practitioners, this regulatory posture means that AI deployment is governed primarily by internal risk appetite and governance capacity rather than by external mandates. Institutions that have invested in robust model governance, clear escalation procedures and documented supervisory protocols tend to be best positioned to deploy AI tools confidently – knowing they can demonstrate responsible use to supervisors when questions arise. The governance infrastructure itself becomes a competitive asset, enabling faster deployment of new capabilities while maintaining the auditability and explainability that regulators expect.

International co-ordination on AI governance in financial services remains fragmented, creating additional complexity for global fund finance platforms. Institutions operating across multiple jurisdictions must navigate varying supervisory expectations regarding explainability, data localisation and human oversight requirements. The Financial Stability Board has published high-level principles for AI in finance, but implementation varies across member jurisdictions. This regulatory fragmentation may increase the value of flexible governance frameworks that can adapt to evolving requirements without requiring fundamental system redesigns.

Where AI is making a practical difference

A compelling case for AI in fund finance comes not from theoretical promise or vendor demonstrations but from specific deployments already operating in production environments. These examples demonstrate concrete value creation rather than speculative efficiency gains. The following examples illustrate the range of applications already delivering value in fund finance operations.

LPA review and diligence

Leading fund finance teams have built AI diligence tools for LPA review that capture the vast majority of standard data points – reportedly in the range of 85–90% – compressing cycle times well beyond what static checklists achieved. These tools run structured extraction queries against documents and return source-linked answers that lawyers can verify efficiently against the original text. The key insight is that AI does not replace attorney judgement; it restructures how that judgement is deployed, allowing professionals to focus on verification and analysis rather than initial extraction.

The efficiency gains in LPA review can affect deal execution timelines. Fund managers increasingly expect rapid responses to financing requests, and the ability to complete diligence in days rather than weeks can influence an institution’s competitiveness in pursuing mandates. Beyond speed, AI-assisted review may produce more consistent results across different reviewers and time periods, reducing the quality variability that accompanies manual processes under time pressure.

Deal screening and financial modelling

At one private credit firm, deal screening that previously required hours of analyst time has been compressed to minutes, allowing teams to evaluate a significantly larger pipeline without proportional staffing increases. The AI system ingests offering memoranda, financial statements and comparable transaction data, producing structured summaries that analysts can review rapidly. Separately, analysts have used AI to build full cash-flow models in roughly two hours rather than the days typically required through manual spreadsheet construction. These are not marginal improvements – they represent fundamental changes to how quickly capital can be evaluated and deployed.

The quality improvements accompanying these efficiency gains may also be relevant for fund finance practitioners. AI-assisted screening may reduce variability associated with manual review by different analysts across different time pressures. Consistent application of screening criteria may help institutions maintain disciplined credit standards even as deal flow increases. The audit trail created by AI systems may also support regulatory compliance and internal risk management by documenting the analytical basis for credit decisions.

Side letters and institutional memory

AI can now ingest entire side-letter populations – sometimes hundreds of letters per fund – and produce consolidated schedules that feed directly into borrowing-base calculations, identifying most-favoured-nation provisions, transfer restrictions and excuse rights. This process previously required days of manual extraction by junior associates.

Retrieval-augmented generation architecture delivers institutional memory from benchmark repositories and precedent databases without client material being used to train the underlying model, allowing firms to query their collective experience without compromising confidentiality. The architecture is not self-executing, however: the confidentiality position depends on contractual no-training, no-retention and sub-processor commitments from the model provider, and on access controls that respect ethical walls and client-imposed restrictions.

These production deployments share common characteristics that distinguish responsible AI adoption, as follows.

  • They automate extraction and organisation rather than decision-making – AI structures information for human judgement rather than replacing it.
  • They maintain lawyer verification as a required step before outputs become authoritative or client-facing.
  • They deliver measurable economic value – faster cycles, broader coverage, fewer transcription errors – with clear return on investment.
  • They operate within existing governance frameworks rather than requiring new regulatory permissions or novel risk approvals.
  • The common thread across many deployments is careful scope definition at the outset. Institutions that attempt to automate complex judgement tasks before mastering simpler extraction workflows can encounter reliability problems that undermine user confidence and slow adoption. Many implementations follow a deliberate progression from narrow, well-defined tasks towards broader applications as the organisation builds experience and refines its verification protocols.

The deal passport: a new architecture for transaction management

Perhaps one of the more significant concepts emerging in fund finance technology is the “deal passport” – a continuously maintained, lawyer-verified record of a facility’s material terms that travels with the transaction from origination through portfolio management and amendment. Rather than recreating term sheets and checklists at each transaction stage – a process that introduces transcription errors and consumes significant professional time – the deal passport maintains a single verified source of truth that accumulates institutional knowledge over the facility’s entire life.

The economic case for the deal passport may increase as institutions’ fund finance portfolios grow in size and complexity. Manual approaches to term management scale poorly – each additional facility adds incremental extraction and verification work, while the risk of error accumulates across the portfolio. The deal passport architecture inverts this dynamic: initial setup requires investment, but each subsequent transaction benefits from established infrastructure, verified precedents and accumulated institutional knowledge.

The architecture operates across three distinct functional layers, as follows.

  • Data layer: stores verified extraction outputs, field definitions, institutional lending policies and precedent libraries. This is the foundation – structured, searchable, version-controlled and auditable, with every data point traced to a source document and a verification event.
  • Reasoning layer: governs how AI tools interact with transaction data through verification queues, playbook criteria, escalation protocols and mandatory attorney sign-off before outputs become authoritative. This layer ensures that speed does not compromise accuracy and that institutional risk appetite is encoded into system behaviour.
  • Execution layer: converts verified legal outputs into structured data that integrates with bank systems for portfolio monitoring, covenant tracking, risk reporting and regulatory compliance. This is where legal work product becomes operational intelligence that the institution can systematically act upon.

The practical benefits of this architecture are substantial and compound over time, as follows.

  • Reduced cycle times for new originations and amendments, as prior verified data carries forward.
  • Reduced transcription error across the documentation life cycle, since verified outputs propagate automatically rather than being manually re-entered.
  • Faster amendment processing when terms change, because the system identifies exactly which downstream data points are affected.
  • Portfolio-level risk visibility that was previously impossible without extensive manual review, enabling identification of concentration risks and covenant patterns across the entire book.
  • Enhanced portfolio-level benefits may be particularly relevant for institutions managing substantial fund finance books. Traditional approaches provide transaction-by-transaction visibility but limited ability to identify patterns, concentrations or emerging risks across the portfolio. The deal passport architecture enables systematic analysis that was previously impractical: identifying which facilities share similar covenant structures, which borrowers have overlapping investor bases and how market standard terms are evolving across recent originations.

Adoption follows a staged path that allows institutions to build confidence incrementally. Institutions typically begin by establishing a data foundation – standardising extraction fields and building verified repositories. They then introduce AI-assisted diligence with human verification, connect outputs to bank systems and ultimately enable supervised automated workflows for routine tasks. Each stage builds on demonstrated reliability before the institution advances. This phased approach is essential because reliability compounds: if an AI system performs at 95% accuracy per step, a five-step automated workflow produces correct results only 77% of the time.

The long-term aspiration extends beyond individual transactions to a centralised workflow platform linking legal AI, bank operational AI and administrative-agent systems. Agentic AI – systems that can execute multistep tasks with limited human intervention – represents the frontier, but current reliability constraints mean that fully autonomous legal workflows remain aspirational rather than operational. The intermediate step is supervised autonomy: AI handles execution while humans handle exceptions, judgement calls and quality assurance at defined checkpoints.

Implementation outcomes may depend in part on change management and user adoption strategies. Organisations deploying AI systems may provide training, establish guidance regarding the role of human judgement in reviewing system outputs, and maintain processes through which users can identify potential errors and suggest improvements. Human oversight is expected to remain relevant as automation capabilities expand.

Risks and controls

The risks of AI adoption in fund finance are manageable with proper governance, training and system design. Courts worldwide have imposed escalating sanctions for AI-fabricated material in legal proceedings, with more than a thousand reported decisions addressing this issue. The penalties have grown more severe as judicial awareness has increased, making quality control an important component of risk management for the deployment of AI in legal work.

For diligence systems, the primary risk is omission – false negatives where material terms are missed – rather than fabrication of non-existent provisions. A system that confidently reports “no change-of-control provision” when one exists buried in a side letter creates greater exposure than one that occasionally identifies a provision that does not exist. Institutions are encouraged to calibrate verification protocols specifically to catch omissions, which would entail different testing approaches than those designed to detect hallucination or fabrication.

Additional risk categories demand careful attention from deploying institutions:

  • privilege and confidentiality risks when client data enters third-party AI platforms without appropriate contractual safeguards, technical controls or informed client consent;
  • cross-border data privacy considerations when investor information from multiple jurisdictions traverses international boundaries through AI processing systems, potentially triggering regulatory obligations in multiple countries; and
  • the emergence of AI-drafted incoming documents from sponsors and fund managers, which may contain characteristic inconsistencies that lenders should train their teams to identify – unusual provision placement, internally contradictory definitions or subtle departures from market standard terms.

Education and structured inputs close the gap between fast adoption and safe adoption. Firms that invest in comprehensive training, obtain informed client consent for AI-assisted work and maintain rigorous supervisory protocols can manage these risks effectively while capturing the substantial efficiency benefits that responsible AI deployment offers.

From a practical standpoint, institutions considering AI deployment in fund finance are also encouraged to evaluate their internal data infrastructure and document management systems. AI tools perform best when they can access well-organised repositories of precedent documentation, standardised templates and cleanly structured historical data. Institutions that have invested in data hygiene and document management will find AI deployment significantly easier and more productive than those working from fragmented, inconsistently organised archives.

Testing and validation protocols may need to evolve alongside AI capabilities. Traditional software-testing approaches may not address all aspects of AI systems, whose outputs can vary based on subtle differences in input phrasing or document formatting. Validation frameworks may incorporate techniques such as adversarial testing, edge-case analysis and ongoing performance monitoring against production data. The objective is generally to assess reliability and performance within documented parameters while recognising the limitations of current AI systems.

Vendor due diligence may present particular challenges in the current AI market environment. The rapid pace of technological change means that vendor capabilities, security practices and financial stability can shift quickly. Institutions deploying third-party AI tools may consider establishing ongoing monitoring protocols in addition to initial due diligence assessments. Contract provisions addressing data handling, service continuity and intellectual property rights are likely to require careful attention given the novelty of many AI deployment scenarios.

Conclusion

AI is not a future consideration for fund finance – it is a present reality reshaping how transactions are originated, documented, managed and monitored. The institutions that may be best positioned to lead are those combining technological capability with governance maturity, recognising that the absence of prescriptive regulation is not an absence of expectation from supervisors or clients.

The opportunity is substantial: faster deal execution, more consistent documentation, portfolio-level visibility that was previously impractical and more efficient allocation of legal experience towards work that genuinely requires human judgement. The challenge is building systems that deliver these benefits reliably, transparently and within a regulatory environment that is still actively evolving.

For lender’s counsel, the key takeaway is that AI may be used as a tool to support more consistent work while maintaining professional judgement and supervisory rigour. The deal passport concept, the evolution of governance frameworks and the regulatory expectation of institutional maturity all point in the same direction – towards workflows that may increase speed, transparency and consistency while maintaining appropriate human oversight. Fund finance’s document-intensive nature may make it a useful environment for evaluating AI-assisted legal workflows, and developments in the sector may provide insights relevant to broader finance practices.

DLA Piper

1251 Avenue of the Americas
New York
NY 10020-1104
United States

+1 212 335 4500

info@dlapiper.com www.dlapiper.com/en-us
Author Business Card

Law and Practice

Authors



A&O Shearman is a global law firm with nearly 4,000 lawyers worldwide. A&O Shearman has one of the largest and most international teams of banking and finance lawyers of any global law firm and one of the leading debt finance practices in the world, known for ground-breaking deals and structures that eventually become market standards. The firm provides clients, including major banks, private credit funds and financial sponsors, with a full-service offering for a broad range of debt finance products for syndicated and private credit transactions, including senior, second-lien and asset-based credit facilities, mezzanine and holdco debt, recurring revenue financings, bridge and bank/bond financings, high-yield bond offerings, securitisation take-outs, debtor-in-possession and exit financings and restructurings. The firm’s global team spans major financial centres, including New York, London, Paris, Amsterdam, Frankfurt, Milan, Madrid, Luxembourg, Singapore, Hong Kong, and Sydney, creating a cross-border network to support its clients’ success.

Trends and Developments

Authors



DLA Piper is a leading global law firm with more than 5,500 lawyers across over 90 offices in more than 40 countries, providing clients with seamless legal services across jurisdictions and industries. The firm advises multinational corporations, financial institutions, private equity sponsors, investment funds, and emerging businesses on their most significant transactions, disputes, regulatory matters, and strategic initiatives. Recognised for its deep local market knowledge and integrated global platform, DLA Piper delivers co-ordinated, cross-border solutions designed to help clients navigate growth, transformation, and complex commercial challenges. The firm is committed to innovation in legal service delivery, leveraging technology, collaboration, and efficient staffing models to provide consistent value and exceptional client service. DLA Piper also maintains a strong commitment to pro bono work, sustainability, diversity and inclusion, and community engagement. Through its collaborative approach and global reach, the firm serves as a trusted advisor to clients operating in an increasingly interconnected business environment.

Compare law and practice by selecting locations and topic(s)

{{searchBoxHeader}}

Select Topic(s)

loading ...
{{topic.title}}

Please select at least one chapter and one topic to use the compare functionality.