Introduction

There is no set formula for “crisis management” – responding to client needs in a crisis often requires a bespoke approach to the specific issue at hand. This necessitates the ability to quickly develop strategies across multiple fronts, often from both legal and public relations perspectives. The areas of law at issue are usually complex and in flux, often overlapping between civil and criminal matters. This is especially true in New York, a perennial hub for finance, business, entertainment, and technology, and, thus, crisis management strategies and legal developments.

For a client facing a crisis, two concepts are often critical:

• maintaining necessary privacy and privilege; and
• protecting a reputation.

These two aspects of a crisis management strategy will usually permeate any overarching legal and public relations plan for addressing the client’s unique issue, whether the client is a multi-national corporation or a private individual.

The Attorney-Client Privilege in the Age of AI

Clients are often comforted by the concept of the attorney-client privilege, and for good reason: candid conversations between a client and an attorney facilitate the interests of justice and have long been a bedrock of our system. However, as a recent opinion by Judge Jed Rakoff in the Southern District of New York demonstrates, clients facing a crisis (or any litigation) should take time to fully understand the limits of the privilege, particularly in this age of prevalent use of generative AI platforms. What may feel like a private “discussion” between a client and an AI chatbot, even about a legal issue the client is facing, is likely to not only be memorialised but also subject to (potentially public) disclosure.

In US v Heppner, 2026 WL 436479 (SDNY 2026), the defendant pleaded not guilty to charges of securities fraud, wire fraud, conspiracy, making false statements to auditors and falsifying corporate records. Defendant asserted that documents he created using a publicly available generative AI platform were protected from disclosure under both the attorney-client privilege and the work product doctrine. The court disagreed on both fronts.

In addressing defendant’s attorney-client privilege arguments, the court first noted that the privilege is construed narrowly because it operates as an exception to the rule that “all relevant proof is essential” for a complete record and for “confidence in the fair administration of justice” (idem). The court ultimately determined that the AI-generated document at issue lacked at least two, and possibly all three, elements of the attorney-client privilege, ie, communications:

• between a client and his or her attorney;
• that are intended to be and, in fact, were kept confidential;
• for the purpose of obtaining or providing legal advice.

Notably, the court considered AI inputs more similar to “communications” than the use of other internet-based software (which, even if not communications, would not be inherently privileged) and all recognised privileges require, among other things, a trusting human relationship. The court found that no such relationship could possibly exist between the AI user and the AI platform.

The court also found that the AI documents were not confidential, given the disclosures about potential use in the AI platform’s terms and conditions. Finally, the court found that the defendant did not communicate with the AI platform for the purpose of obtaining legal advice, as the communications were not made at counsel’s suggestion or direction. Interestingly, the court noted that if counsel had directed defendant to run the AI searches, the platform “might arguably be said to have functioned in a manner akin to a highly trained professional who may act as a lawyer’s agent within the protection of the attorney-client privilege.” This analysis potentially opens the door to some form of privilege protection for communications between clients and an AI platform, even without direct involvement by counsel, when those communications are done at counsel’s direction.

In determining that the AI documents were likewise not protected from disclosure under the work-product doctrine, the court found the facts that they were not “prepared by or at the behest of counsel” and did not “reflect defence counsel’s strategy” dispositive. The court rejected the holding of an earlier Southern District of New York decision that concluded that the work product doctrine is not limited to materials prepared by or at the direction of an attorney. Judge Rakoff noted long-standing Second Circuit precedent stressing that the purpose of the work-product doctrine is to “protect lawyers’ mental processes.” As there was no dispute that the defendant acted on his own when creating the AI documents, they were not protected.

Clients facing a crisis should exercise extreme caution in “communicating” with generative AI platforms, particularly when done without the direction of counsel. While this area of the law is certainly dynamic and developing, based on the Heppner decision, courts in New York have shown that long-standing privilege and work-product principles apply to use of this emerging technology and what a client might have thought was a private “communication” with AI could end up in the hands of the opposition and even show up in a public court filing.

Reputation Management Litigation

Clients facing a crisis may seek to manage their reputation through litigation. New York benefits from a robust and well-developed body of law in this area. Clients should, however, be aware that recent trends favour strengthening protections for the First Amendment and public participation rights.

Particularly in this age of social media, potentially harmful and defamatory information can spread like wildfire and have a negative impact on a person’s or a business’s reputation overnight or even more quickly. Clients facing a crisis should understand the benefits and the potential limits of litigation as a reputation management strategy.

New York Civil Rights Law § 74 – fair reporting

Generally, New York Civil Rights Law § 74 provides immunity for fair and true reports of judicial proceedings.

One of the purposes of New York Civil Rights Law § 74 was to “encourage accountability through public legal proceedings and the truthful reporting of whatever transpires in open court.” Rapaport v Finkelman, 2025 WL 2494315, *5 (S.D.N.Y. 29 August 2025) (citing Gannett Co. v De Pasquale, 43 N.Y.2d 370, 378 (1977)). The key question in determining whether a statement is immune is whether an ordinary viewer can, from the publication standing alone, determine that it reports on a judicial proceeding.

In a recent decision dismissing defamation claims brought by former NYU student Gideon Rapaport against a Twitter user, the Court focused on immunity under New York Civil Rights Law § 74’s fair report privilege as well as the distinction between non-actionable opinions and false statements of fact. Finkelman, 2025 WL 2494315, *5.

The complaint involved the defendant’s tweet of a photograph plaintiff alleged was “fake”, which was in the style of a “wanted” or “public enemy” poster. The photograph was the subject of ongoing litigation between the plaintiff and his former classmates in a separate case. The photograph included the words “Do not admit” along with the plaintiff’s name and the law firm name “Kirkland and Ellis.” Plaintiff alleged that the photograph “taken together with the context in which it appears” was susceptible to a defamatory connotation in that it allegedly implied that plaintiff was a “trespasser” and that he was “unfit to practice his profession.”

In evaluating whether the post in Finkelman was defamatory, the Court emphasised that allegedly defamatory statements must be reviewed against the “whole apparent scope and intent of the writing” rather than in isolation. Finkelman at *6 (quoting Kesner v Dow Jones & Company, Inc., 515 F.Supp.3d 149, 170 (S.D.N.Y. 2021). The court focused on whether the publisher wrote or suggested anything that would cause an ordinary reader to have a different opinion than one they would have if they had merely read the complaint about which the publication was made. The court answered this question in the negative, concluding that the posts at issue simply reported on the ongoing underlying case.

The photograph posted by the defendant in Finkelman was not attached to the complaint in the underlying case but instead received by an “anonymous” tipster. The court found this fact insignificant, noting that information about a particular judicial proceeding is derived from secondary sources does not mean that New York Civil Rights Law § 74 is inapplicable.

Finkelman illustrates the latitude defamation defendants are afforded when commenting on ongoing judicial proceedings, even in posts that include information from sources other than the proceedings themselves.

Defamatory Connotation and Statements of Opinion

It is often colloquially said that a statement is immune to a claim for defamation so long as it is an expression of opinion. While this may generally be an accurate summary, what qualifies as a pure statement of opinion may not be so obvious. A recent New York case that illustrates the line between pure opinion and actionable statements of fact is Fradkoff v Winston. There, the Southern District of New York considered when statements that may facially seem like opinion cross the line into actionable defamatory statements. That case dealt with a plaintiff and a defendant who were both employed by a prestigious jeweller. Defendant took control of the company following his father’s (the company’s founder) death. Defendant published a book about his father and his leadership of the company, which portrayed the plaintiff negatively. Plaintiff sued for defamation.

Plaintiff highlighted five particular statements within the book he alleged were defamatory, and the court considered the context of the book as a whole in evaluating the defamation claim. The five statements illustrate how New York courts draw the line between non-actionable opinion and actionable defamatory statements. They were:

• “I sensed they were stealing my family’s money”;
• “it takes a thief to know a thief”;
• “because there was honour among thieves, Fradkoff never called Bochatay on any of his suspected transgressions”;
• “Fradkoff… was betting the farm using Harry Winston’s funds… sensing not only a high crime but a major disaster in the making”; and
• ”Serge [Fradkoff] paid us back in full, reversing his thievery.”

In adjudicating the defendant’s motion to dismiss, the Court determined that the fourth and fifth statements were not protected opinion and therefore the complaint could proceed on those statements. It was determined that the first, second and third statements, on the other hand, were non-actionable.

Perhaps most interestingly, the Court determined that the second and third statements were so vague and hyperbolic that they merely constituted an opinion rather than a statement of fact. The court rejected the plaintiff’s argument that the statements constituted a statement of fact about the plaintiff’s criminal actions, noting that “there is honour among thieves” and “it takes a thief to catch a thief” are generally understood as “idiomatic expressions or rhetorical flourishes, rather than direct factual accusations.” The Court cited Jewell v NYP Holdings, Inc., 23 F.Supp.2d 348, 382 (S.D.N.Y. 1998) in reaching this conclusion, quoting the portion of that opinion that said: “[s]ome language, regardless of context, is so vague and hyperbolic that it can only reasonably be viewed as language of opinion.”

On the other hand, despite the use of the word “sensing” like in statement one, the Court found that the fourth statement was actionable. The fourth statement was preceded by a reference to a specific circumstance where the defendant allegedly faced the threat of “exposure to the French authorities” due to alleged “misuse of company funds” and a subsequent repayment. In finding that the statement was actionable, the Court noted that the language “high crime in the making” had a “more precise” meaning than the other, non-actionable statements because, given its context, it identified a specific instant where Fradkoff allegedly committed misconduct. Most importantly, the context of the statement in the book at issue did not provide readers with sufficient facts that supported the allegation. Therefore, the Court found, this lack of explanation implied the existence of “undisclosed facts” which caused “readers… to assume those facts are unfavourable to the statement’s subject” idem at *5 (citing Coleman v Grand, 523 F. Supp. 3d 244, 262 (E.D.N.Y. 2021)).

Anti-SLAPP

In November 2020, New York expanded First Amendment protections under New York’s anti-SLAPP (Strategic Lawsuit Against Public Participation) statute (N.Y. Civ. Rts. L. § 76-a). Before the 2020 amendment, New York’s anti-SLAPP law generally applied only to suits arising from a public application or permit. The expanded statute applies to any action based on a communication in a “public forum in connection with an issue of public interest” as well as “any other lawful conduct in furtherance of the exercise of the constitutional right of free speech in connection with an issue of public interest, or in furtherance of the exercise of the constitutional right of petition.”

Regarding application to suits arising from communications in a public forum about an issue of public interest, the statute generally provides that a plaintiff must make a showing “by clear and convincing evidence” that the alleged tortfeasor acted with actual malice, ie, “with knowledge [of the statement’s] falsity or with reckless disregard of whether it was false…” idem § 76-a(a). The statute defines “public interest” broadly to include “any subject other than a purely private matter.”

Interpretation of this statute has been steadily developing in New York courts since the amendment. Anti-SLAPP motion practice has effectively evolved into mini-summary judgment proceedings, typically forcing plaintiffs to produce important evidence before discovery commences. Specifically, New York courts have recently clarified the standard an opposition must meet to withstand dismissal on an anti-SLAPP motion under CPLR 3211(g)(1). Pursuant to the CPLR, the opposition must demonstrate that the claim at issue has a “substantial basis.” The “substantial basis” standard was analysed in Reeves v Associated Newspapers, Ltd., 232 A.D.3d 10, 218 N.Y.S.3d 19 [1st Dept. 2024]. There, the First Department reasoned that “substantial basis” under New York’s anti-SLAPP law means “such relevant proof as a reasonable mind may accept as adequate to support a conclusion or ultimate fact” (Reeves at 22). The Reeves court went on to note that this is the same “substantial evidence” standard that “has been equated with the ordinary summary judgment standard”, which requires the submission of evidence such as an affidavit rather than simply relying on allegations in the complaint (idem at 24-25).

New York’s anti-SLAPP jurisprudence has also made clear that social media platforms generally qualify as “public forums,” triggering anti-SLAPP applicability. The consequence is that claims that on their face may appear to be private defamation actions could face anti-SLAPP scrutiny when a social media platform is involved (see, eg, Nelson v Ardrey, 231 A.D.3d 179, 216 N.Y.S.3d 646 (2nd Dep’t 2024)).

Given these developments in New York’s anti-SLAPP law, online posts are likely to satisfy the “public forum” prong of the anti-SLAPP statute. However, to be subject to the statute’s protections, the statement must still be one related to a matter of public interest. As recent New York decisions make clear, that definition may not be as narrow as it initially appears. For example, it has become clear that New York courts generally consider online reviews matters of public interest. Reviews have been characterised as “consumer advice” concerning service providers (Warren v Gottlieb, 2025 WL 834463, **2-3 (N.Y. Sup., New York Cnty. 14 March 2025) (collecting cases)).

Additional efforts are underway in the New York legislature to further expand the anti-SLAPP statute. Senate Bill S4282 proposes an amendment to the CPLR that would automatically stay discovery pending an adjudication of an anti-SLAPP motion. This further demonstrates the New York trend toward an early merits determination where a suit raises potential anti-SLAPP concerns.

As part of a crisis management strategy that may involve bringing a defamation claim to safeguard a reputation, a client should be prepared to lay out evidence sufficient to meet the “substantial evidence” standard to survive an anti-SLAPP motion early in the case, and potentially without the benefit of first obtaining discovery.