“AI Washing” – The New Deceptive Marketing Technique
Artificial intelligence, or AI—technology that enables computers to mimic human learning, comprehension, problem solving, and/or decision making—is the “new” hot thing. The trend is driven by innovation in so-called “generative AI”: text, image, audio, and video generating computer programs able to utilize vast data sets to create realistic looking and sounding output. Many imagine that AI first stormed the world in November 2022 with the launch of ChatGPT, but AI tools have been around for years. IBM’s Watson AI system crushed its human competitors on Jeopardy! over thirteen years ago. But with this latest generation of generative AI, every company wants a role in the AI revolution. With this increased desirability, it seems that suddenly everything is being advertised as being powered by AI, even when it isn’t.
AI hype has led to a deceptive marketing practice called “AI washing.” Similar to “green washing” – exaggerating positive environmental impact to distract the public or investors from less flattering news – AI washing exaggerates a company’s effective use of AI technology in its products or operations. AI washing is intended to make a company appear more sophisticated or technologically advanced than it really is by linking it to the technological trend of the moment.
AI washing can take different forms. A company could outright lie about the existence of AI in its products, exaggerate AI’s impact on the business or its capabilities, or falsely suggest a new AI system can out-perform existing, non-AI products or systems. AI washing can also involve exaggerating a new AI operation’s sustainability or suggesting that AI represents a new direction when its use in a given situation is just a gimmick. AI technology requires massive computing power and data storage as well as up-to-date data sets to feed the algorithm, all of which are costly. It is equally AI washing for a company to tout its new AI system without acknowledging those costs.
Even the largest corporations engage in practices that could be considered AI washing. Amazon has faced accusations of AI washing for its Just Walk Out Program. For Just Walk Out, Amazon publicized an AI-powered system that allowed Amazon Fresh and Amazon Go shoppers to pick up their items and leave without paying because the AI-backed sensors would identify items chosen and bill customers automatically. Pitching the service as an AI play raised Just Walk Out’s profile, but also hid the fact, reported earlier this year, that Amazon needed around 1,000 workers in India to manually check almost three-quarters of transactions. Amazon denied the reports but did admit that the Indian workers were reviewing the system. The takeaway is that Just Walk Out used AI, but not as effectively as Amazon suggested.
The SEC Moves to Clean Up AI Fraud
The Securities and Exchange Commission has taken action on AI washing in moves that portend future scrutiny of the practice. In March 2024, the SEC announced it had settled charges against two investment advisors, Delphia (USA), Inc. and Global Predictions, Inc., for making false and misleading statements about their purported use of artificial intelligence. The firms settled for $225,000 and $175,000 respectively. According to the SEC, both firms were marketing to clients that they were using AI in ways they were not. SEC Chair Gary Gensler warned about false claims concerning AI use. In the press release announcing the settlement, the outgoing Director of SEC’s Division of Enforcement, Gurbir Grewal, while indicating that the Commission would be focused on AI washing in the financial services industry, stated, “[W]e are committed to protecting [investors] against those engaged in ‘AI washing.’”
In June 2024, the SEC followed up with an AI washing charge against an issuer, charging the CEO and founder of AI startup Joonko Diversity, Inc., Ilit Raz, with defrauding investors. According to the complaint, Ms. Raz marketed Joonko as a technology platform using AI to match customer firms with diverse job candidates to help clients achieve Diversity, Equity, and Inclusion goals, when in reality the platform did not function as Ms. Raz claimed at all. In a press release announcing the complaint, Mr. Grewal called the case “an old school fraud using new school buzzwords like artificial intelligence and automation.” He promised continued policing “against AI-washing and the type of misconduct alleged in today’s complaint.”
Chairman Gensler has commented on this topic before. In prepared remarks to Yale Law School in February 2024 on AI more generally, the SEC Chair called out the potential for AI washing and offered advice to those making public statements: “If a company is raising money from the public, though, it needs to be truthful about its use of AI and associated risk.” He continued, “AI washing, whether it’s by companies raising money or financial intermediaries, such as investment advisers and broker-dealers, may violate the securities laws.” Recently, on an episode of his YouTube show “Office Hours,” Mr. Gensler once again warned about AI washing, signaling a further crackdown. He broadened the focus to discuss registrant filings, explaining that companies should consider whether AI discussions might be material to investors. Clearly, the SEC will stay focused on this problem.
Investors Won’t Be Left Out to Dry
In addition to SEC scrutiny, AI washing is already generating private securities fraud cases under the Exchange Act and Securities Act. Indeed, in 2024, investors brought at least eight prominent AI washing cases alleging Exchange Act claims.
One prominent example is D’Agostino v. Innodata Inc., et al., No. 2:24-cv-00971 (D.N.J.), where shareholders of Innodata Inc., a global data engineering company focusing on AI, allege that Innodata made false claims about its “proprietary, state-of-the-art” “core AI technology stack,” when in reality Innodata lacked viable AI technology, its platform was a rudimentary software, Innodata was not going to meaningfully use AI for new contracts, and it was not effectively investing in research and development for AI. According to the plaintiffs in the Innodata suit, the truth emerged in a short seller report exposing the company’s AI washing practices.
In Hoare v. Oddity Tech Ltd. et al., No. 1:24-cv-06571 (S.D.N.Y.), shareholders of Oddity, a consumer tech platform that uses AI to identify customer needs and solutions in the beauty and wellness products space, allege that the company overstated its AI technology and capabilities, and the extent to which that tech-nology drove sales. According to the suit, the truth emerged in a short seller report describing discussions with former employees who called the company’s AI “nothing but a questionnaire.”
More litigation seems inevitable as public companies rush to tout their ultimately ephemeral AI-integration and products. Further suits could be driven by companies that include AI disclosures in their public filings without adequately disclosing risk, thus creating the impression of AI usage that materially differs from reality. Chairman Gensler has already hinted that generic risk disclosures for AI will not suffice. In his prepared remarks to Yale, he said “investors benefit from disclosures particularized to the company, not from boilerplate language.”
Section 11 claims under the Securities Act for AI washing in a registration statement may be even more common than Exchange Act Section 10(b) claims. For any startup, the need to appear cutting-edge is powerful, and companies may be tempted to exaggerate AI tools that are actually incidental to the product. Additionally, legitimate technology startups trying to innovate in AI may misstate their readiness to incorporate the tools or overstate what they can do. This is already a problem. According to a 2019 report by tech investment firm, MMC Ventures, 40% of new tech firms that described themselves as “AI start-ups” in fact used virtually no AI at all.
AI washing may be an emerging problem for investors, but as its moniker suggests, it is just another form of misleading customers and investors: an old school fraud using new school buzzwords. The securities laws apply to the new as well as the old.