Is that company powered by proprietary AI insights and solutions or did it just spend some time futzing around with ChatGPT?
Sorting out the reality behind vague buzzwords has become a key skill in the era of generative AI hype, in which a sense of FOMO has pressured startups and big corporations alike to perhaps overstate their AI capabilities. But on top of losing trust with consumers or investors if these claims don’t pan out, such inflated language—dubbed “AI washing”—can come with legal risks.
Even as Wall Street investors worry about an AI bubble, businesses are still gunning to show off AI bona fides. Startups centered on AI may be more likely to nab funding, and S&P 500 companies that mention AI in earnings calls, which reached an all-time high in 2023, tend to see better stock performance on average.
‘New school’ fraud
But the Securities and Exchange Commission (SEC) has begun to crack down on companies that make false claims about AI in recent months following a series of warnings to publicly traded companies from Chair Gary Gensler. In June, for instance, the commission charged the CEO and founder of a now-shuttered recruitment platform called Joonko in “an old-school fraud using new-school buzzwords like ‘artificial intelligence’ and ‘automation,’” as Gurbir Grewal, director of the SEC’s Enforcement Division, said in a statement.
“As more and more people seek out AI-related investment opportunities, we will continue to police the markets against AI washing,” Grewal added. “It is critical for investors to beware of companies exploiting the fanfare around artificial intelligence to raise funds.”
Beyond enforcement actions, companies also face a risk of shareholder lawsuits, much like with greenwashing or, in rarer cases, “cloud-washing” according to David Shargel, a partner at law firm Bracewell.
One such suit filed last month targeted a makeup company called Oddity for violating US securities law in misleading investors. Plaintiffs accused the company of having “overstated its AI technology and capabilities, and/or the extent to which this technology drove the
company’s sales.” The company, which claims to use AI to develop and recommend products, did not immediately respond to a request for comment.
Muddled meaning
The concept of AI washing has been around since well before ChatGPT debuted in 2022, according to Kjell Carlsson, head of AI strategy at Domino Data Lab. Previous waves of inflated claims have emerged around predictive analytics and other machine learning technologies, he said. It doesn’t help that AI is a vague term that has come to encompass a whole set of different fields of tech, multiple experts we talked to said.
Keep up with the innovative tech transforming business
Tech Brew keeps business leaders up-to-date on the latest innovations, automation advances, policy shifts, and more, so they can make informed decisions about tech.
Carlsson said the pressure puts vendors in a tough spot. “It’s very much a damned if you do, damned if you don’t kind of a thing…where you kind of have to say that you have AI capabilities, because everybody else says that they have AI capabilities,” he said. “And if you don’t, folks are like, ‘Well, why don’t you?’”
Akhilesh Agarwal, COO and EVP at ApexAnalytix, said there can also be disconnects between marketing and technical departments.
“Marketing will be incentivized to say, ‘We’ve got everything right now,’” he said. “Engineers are on the other end of the spectrum and saying, ‘We have a lot of work to do, a lot of experimentation to do, a lot of learning to do before we can launch a product.’”
Let’s be clear
Laura Lin, a partner at Simpson, Thacher & Bartlett, said beyond striving for basic accuracy in AI claims, companies should steer clear of describing in-house versions of open-source tech as “proprietary” and delineate between what’s currently possible and predictions about the future.
“It is also equally important to refrain from implying that a company is using a ‘secret AI sauce’ in-house when there may not be one,” Lin wrote in an email. “In the same vein, they should be mindful of how they describe what AI is currently doing for them and not blurring the lines between what is possible in the present and what they anticipate AI can do for them in the future.”
As some companies struggle to break out of the experimentation phase, however, generative AI may be losing some of its initial luster. Research firm Gartner has officially moved the trend into the “trough of disillusionment” on its hype cycle map, the phase where an emerging tech inevitably fails to live up to outsized initial expectations.
Does this mean that the pressure to AI wash could die down a bit, too? “I hope so,” Gartner VP analyst Nicole Greene said.
“These exaggerated claims then leave organizations feeling very disillusioned by the technology, when really it might not be the technology that didn’t serve its purpose,” Greene said. “It just might have been a marketing claim that didn’t meet the expectation.”