Yes. In 2026, federal prosecutors pursued criminal fraud charges in cases where founders made materially false claims about their AI capabilities to investors — going beyond the civil SEC enforcement actions that characterized earlier AI washing cases. The prosecution of companies like Nate for claiming AI automation that was actually human labor signals that the government treats investor-facing AI misrepresentations as potential wire fraud, not just securities violations.
AI washing — claiming AI capabilities that a company's product does not actually have, or obscuring the extent to which marketed "AI" functions involve human labor — had been prevalent enough by 2023 to generate SEC guidance and a handful of civil enforcement actions. The shift from civil to criminal enforcement in the Nate prosecution reflects a specific prosecutorial calculus: when misrepresentation involves material facts upon which investors made capital allocation decisions, it satisfies the elements of wire fraud even without securities law violations. The Nate allegations were that the company's checkout automation product, marketed to investors as AI-driven, was substantially operated by human contractors in the Philippines, and that this reality was concealed from investors who funded the company on the basis of AI capability claims. The criminal charge does not require proving intent to defraud through complex securities law analysis — wire fraud under 18 U.S.C. §1343 requires a intent to defraud, use of wire communications, and a scheme involving material misrepresentation. Investor pitch decks, email communications, and demo materials sent over the internet all satisfy the wire element. For boards and investors reviewing AI claims in materials going to new investors, the relevant question is no longer just SEC exposure — it is whether the claims could form the basis of a wire fraud allegation if the underlying capabilities are later found to be materially different from what was represented.
Academic fraud research consistently identifies concentrated founder control without independent oversight, weak board governance, pressure to hit investor-promised growth metrics, and rapid fundraising without adequate financial controls as the leading predictors of startup fraud. Companies where the CEO also controls the board, lacks a qualified CFO, and has raised money on aggressive forward-looking projections are statistically most vulnerable.
The empirical study of 614 venture-backed startup fraud cases — the largest such analysis I am aware of — produces findings that should materially affect how investors think about governance structure at portfolio companies. The headline finding is that founder-controlled boards are associated with an 88% higher likelihood of fraud compared to companies with more balanced governance. More important than the magnitude is the analytical mechanism: the study found that governance variables — board composition, independent director presence, investor board representation — predicted fraud incidence more strongly than founder personal characteristics. This is analytically significant because it suggests that fraud in VC-backed companies is not primarily a function of recruiting the wrong founders. It is partly a function of creating the structural conditions in which misconduct is more likely to go unchecked. Investors who require or accept founder-dominated governance structures without corresponding oversight mechanisms are, on average, accepting higher fraud risk. The countervailing argument — that founder control produces better company performance and more decisive decision-making — may be correct on those dimensions. The data suggest it also produces higher fraud incidence. The practical resolution is not to eliminate founder governance but to ensure that the oversight mechanisms accompanying it — independent directors with real authority, audit functions appropriate to stage, investor board members with actual visibility into operations — are sufficient to function as a check. A governance structure in which the only person with full visibility into financial claims is also the person making those claims to investors is a structure that the data say creates risk.
The Lux Optics litigation established that spoliation sanctions — penalties for destroying evidence — can be outcome-determinative in startup disputes. Even routine deletion of messages or emails after a lawsuit is reasonably anticipated can result in adverse inference instructions to the jury, meaning courts can assume the destroyed evidence was damaging. Founders facing any dispute should issue a litigation hold immediately and stop deleting communications.
Lux Optics v. de With, one of the significant co-founder disputes of 2025–2026, illustrates the escalation pattern that has become more common as VC litigation rises. The case involved allegations of data destruction — approximately 500 gigabytes of company data allegedly deleted by the departing co-founder in the period surrounding his departure — along with disputed IP assignments and third-party relationships involving Apple. The data deletion allegation is notable not because it is rare in co-founder disputes, but because it is nearly always counterproductive for the party who does it. Spoliation of evidence — destroying documents or data that a party knows or should know will be relevant to litigation — creates independent legal liability, generates adverse inference instructions at trial, and eliminates any negotiating advantage the departing founder might have had. Courts, in my experience, are not sympathetic to the data-deletion position regardless of the underlying merits of the dispute. More broadly, the Lux Optics case typifies what has changed about co-founder disputes: what used to resolve through negotiated separation agreements is increasingly proceeding to formal litigation with escalating legal costs, public records that follow both parties' professional reputations, and outcomes that are less favorable for everyone than a well-negotiated exit would have been. The VC litigation dataset corroborates the anecdotal pattern: fiduciary duty claims — both against founders and against investor-affiliated board members — represent approximately 40% of VC-related lawsuits in the final period of the 2014–2025 study window, up significantly from the earlier period.
Board members face personal liability when they had actual knowledge of misrepresentations and approved or failed to correct investor communications, when they failed to establish adequate oversight mechanisms over management's public statements, or when they received the material benefits of the fraud. The business judgment rule protects good-faith decisions, but it does not protect directors who were deliberately uninformed about practices they had reason to investigate.
The 2025–2026 enforcement cases have so far targeted founders, not their VC-backed board members. That focus should not be read as an indication that investor exposure is absent. Wire fraud conspiracy liability does not require that a co-conspirator personally made the misrepresentation — it requires that they knowingly participated in a scheme to defraud. A venture investor who sits on a board, receives materials claiming AI capabilities, approves fundraising rounds built on those claims, and fails to conduct any independent verification of the technical claims has a different exposure profile than one who documented diligence, raised questions, or required an independent technical review. The inquiry process matters: not because it creates a complete defense, but because it establishes the state of mind element that prosecutors need to prove. For investor board members currently serving at AI startups, the practical response is to ensure that AI capability claims in materials going to new investors or customers have been reviewed by someone with technical competence, that the review is documented, and that any known gap between marketing language and product reality is not actively concealed. The SEC's continued attention to AI disclosure in public company filings — including pre-IPO S-1 registration statements — will apply even greater scrutiny to the same claims in the context of public capital markets transactions, where the regulatory framework for material misrepresentation is more developed and enforcement tools are extensive.
Founders should establish independent audit and compensation committees with genuinely independent directors, maintain board minutes that accurately document decisions and information presented, retain outside counsel to review all investor communications before distribution, and create a disclosure controls policy that requires legal review of any claims about product capabilities. These steps create a record of good-faith governance that is the primary defense in both civil and criminal proceedings.
The fraud data and litigation trends converge on a set of governance practices that are more valuable now than they were in 2022. For founders: AI and technology capability claims in investor materials should accurately reflect current product reality, not aspirational capabilities or human-assisted performance positioned as automated. The gap between what you believe your product will be able to do and what it does today is the gap that prosecutors and enforcement staff have been focusing on. Build a board with at least one independent member who has the technical or financial background to ask hard questions, and give that person real access to information rather than curated management presentations. For investors: the fraud study's implication is that governance structure is a material risk variable, not just a portfolio value-add. A governance review of portfolio companies — focused on whether independent oversight mechanisms are actually functional — is a reasonable fraud risk mitigation measure. For both groups: the co-founder dispute trajectory argues for front-loading governance documentation, including clear descriptions of founder roles, decision-making authority, and the process for resolving disputes, before those questions become contested in a dispute. See the related pieces on co-founder disputes and founder pushouts and conflicts of interest in startup legal representation for the related legal and professional responsibility dimensions.
Gurpreet's view is that AI washing is really just the latest version of a problem that has always existed: founders misrepresenting their company to investors. The enforcement wave does not require a new framework to understand or a new compliance program to address. It requires honesty. When you are talking to investors, tell them the truth — about what the product does today, about what it does not do, about where human labor is involved and where it is not. When you are presenting projections, show your work — the sources, the assumptions, the reasoning that got you from here to there. There is no magic formula for staying out of legal trouble on this. Founders who tell the truth about their companies, even when the truth is less impressive than the narrative they wish they could tell, do not end up in wire fraud prosecutions. It is straightforward.
Gurpreet S. Bal is a Partner at Foley and Lardner LLP in Silicon Valley, where he advises startups, founders, and investors on SAFE financings, venture capital rounds, mergers and acquisitions, acquihires, and IPOs. He has represented clients in hundreds of transactions with aggregate deal value exceeding $60 billion across AI, semiconductors, fintech, and emerging technology.