The Federal Trade Commission issued a strong warning to vendors: stop lying about your AI.
"Some products with AI claims might not even work as advertised in the first place," Michael Atleson, a lawyer with the FTC's Division of Advertising Practices wrote in a blog post..
"Marketers should know that -- for FTC enforcement purposes -- false or unsubstantiated claims about a product's efficacy are our bread and butter," Atleson continued.
Generative AI and AI-washing
"This is one of those places where there's a lot of attention on AI," said Michael Bennett, director of the education curriculum and business lead for responsible AI at Northeastern University. "Companies have been making extraordinary claims, some of them true, some of them false, since companies existed."
Overextended marketing claims are not unique to the U.S. and to the marketing hype that has sprung up around ChatGPT and other new generative AI systems.
A 2019 study that focused on 2,830 startups in 13 European countries marketing themselves as AI found that only 1,580 startups fit the AI definition of "computer systems that can perform tasks that usually require human intelligence."
"AI-washing is rampant," said RPA2AI Research analyst Kashyap Kompella. AI-washing is when vendors lie about a product being AI-enabled or powered.
"AI vendors are able to get away with these exaggerations because artificial intelligence is such a large umbrella term," he continued. "There's a joke that 'when you are fundraising it's AI, when you are hiring it's machine learning, and when you are coding its linear regression.'"
Implications of lying
However, the implications of such lies and exaggerations are not a joke. False claims could lead not only individual vendors but also entire industries to get a bad reputation, Kompella said.
CEOKashyap Kompella, RPA2AI Research
Although it's hard to tell which vendors are simply puffing (overstating their belief on what their product can do so consumers buy the product) or intentionally lying about AI, the FTC appears to be simply trying to fulfill its duties of protecting consumers the way it has always done, Bennett said.
Meanwhile, another federal move to regulate AI technology in the absence of actual legislation to do it came with the White House Office of Science and Technology Policy’s blueprint for the AI Bill of Rights in October, which identifies five principles to consider for those designing AI systems.
"The FTC is carrying a lot of the weight materializing and realizing those principles even if they are not law," Bennett said. "The FTC's blog post seems to be an effort to check a universal phenomenon across a commercial space, including a commercial space in which AI vendors are providing services and goods."