The Federal Trade Commission has filed a complaint against IntelliVision Technologies Corp., a San Jose-based company, for allegedly making false and misleading claims about its AI-powered facial recognition software.
The FTC alleges that IntelliVision falsely advertised its software as being free of gender and racial bias and possessing one of the highest accuracy rates on the market without substantiating these claims.
According to the FTC's complaint, IntelliVision lacked evidence to support its assertions about the software's accuracy and its bias-free performance across different genders, ethnicities, and skin tones.
The complaint also claims that IntelliVision's claim of training its software on millions of faces was misleading, as the company actually used images of approximately 100,000 unique individuals and created variants of these images using technology.
Samuel Levine, director of the FTC’s Bureau of Consumer Protection, emphasized the importance of truthfulness in advertising AI systems, saying, "Companies shouldn’t be touting bias-free artificial intelligence systems unless they can back those claims up."
IntelliVision's facial recognition technology is utilized in home security systems and smart home touch panels sold by Nice North America, LLC. The FTC also challenged IntelliVision's claims about its anti-spoofing technology, which the company asserted could not be tricked by photos or video images, citing inadequate evidence to support this claim.
Proposed consent order
Under a proposed consent order, IntelliVision will be prohibited from misrepresenting the accuracy and efficacy of its facial recognition software. The order also restricts the company from making unsupported claims about the software's performance across different demographics and its ability to detect spoofing. Any future claims must be backed by competent and reliable testing.
This case marks the second major action by the FTC against AI facial recognition technology in the past year. Previously, the FTC settled with Rite Aid, banning the retailer from using facial recognition technology for surveillance due to inadequate consumer protection measures.