Consumer groups urge FTC probe into Grok's spicy mode

If you ask most mainstream AI models to strip down a celebrity, you will get a stern refusal or a quick account ban. Elon Musk’s xAI, on the other hand, decided to take a very different route.

A coalition of 15 consumer protection groups, led by the Consumer Federation of America (CFA) and joined by EPIC and the Tech Oversight Project, just sent an urgent letter to the Federal Trade Commission (FTC) and attorneys general across all 50 US states and DC, demanding an immediate investigation into Grok and its newly released Imagine media generation tool.

The target of their outrage is Grok's controversial "Spicy" mode, which openly nudges users to produce NSFW content.

Unprompted deepfakes and lax barriers

The coalition pointed directly to recent testing by The Verge, where the Imagine tool generated topless deepfakes of pop star Taylor Swift without the reviewer even requesting nudity.

While xAI currently prevents users from uploading real-life photos directly into "Spicy" mode, the groups warned that the software still generates nude clips resembling specific, real individuals. The letter noted that if xAI ever lifts restrictions on user-uploaded images, it would unleash a flood of nonconsensual deepfakes, noting Musk's track record of tearing down guardrails under the banner of free speech.

While distributing AI-generated nudes of real people is illegal under the US Take It Down Act, the groups acknowledged its language might not neatly fit Grok's current setup. Instead, they want the FTC to scrutinize xAI for potential violations of nonconsensual intimate imagery rules.

The letter also took aim at Grok’s age verification. Currently, the only barrier stopping minors from churning out explicit media is a single pop-up asking if they are over 18, with one prompt pre-filling the birth year as 2000. Consumer groups argue this laughably bypassable gate likely violates the Children’s Online Privacy Protection Act (COPPA) as well as state age-verification statutes.

Why it matters

Musk has deliberately positioned xAI as the uncensored, anti-woke counterweight to OpenAI and Google. While that rebel posture drives plenty of buzz, shipping a model that unprompted spits out explicit celebrity deepfakes feels less like frontier innovation and more like reckless product management.

For creators and regular users, dipping into Grok's Spicy mode right now is a legal minefield. We think regulators will have little choice but to step in here. By making age verification essentially an honor system with a pre-filled birth year, xAI gave watchdogs an open-and-shut target under existing child safety laws.

If the FTC or state attorneys general open formal probes, expect this to set the baseline for nonconsensual AI imagery. Platforms flirting with loose NSFW guardrails will likely face mandatory, friction-heavy identity verification, ending the era of turning off safety filters for cheap growth.