An investigation has revealed that Grammarly and Superhuman, two popular productivity platforms, have been using real people's names within their AI-generated suggestions without obtaining consent. The practice, which involves referencing experts and public figures to lend authority to AI outputs, places these companies in a legally ambiguous and ethically questionable position. The core function is simple: a user receives a writing suggestion or a drafted email reply that cites a specific individual's name or implied endorsement. The names are drawn from public sources, but their commercial use within a paid software feature was never authorized by the people involved. There was no prior notification, and until recently, no straightforward way to opt out. For the individuals named, the risks are tangible. An AI tool could misrepresent their views or attach their reputation to flawed advice, damaging their professional standing without their knowledge. Legally, state-based 'right of publicity' laws may offer protection against such unauthorized commercial use, though this specific application remains untested in court. In response to questions, both companies indicated they are developing opt-out procedures. However, this reactive approach places the entire burden on the individuals whose identities are being utilized, not on the firms building the features. The issue arrives as global regulators sharpen their focus on AI transparency. With the EU's AI Act now in effect and similar discussions active in the U.S., the pressure for clear consent frameworks is mounting. For the many corporations that license these tools, this represents a potential liability. Legal and compliance departments are now tasked with understanding how these AI assistants handle third-party identities. As AI becomes a standard feature in workplace software, the industry's habit of moving fast now clashes with a fundamental question: who controls the use of a person's name?

Source: Webpronews