Inside Elon Musk's bet to hook X users that turned Grok into a porn generator; sources say xAI's AI safety team was just two or three people for most of 2025
all in a bid to keep users engaged and away from the AI system's original purpose
Washington Post
Context & Ripple Effects
The report reframes Grok’s sexualized-image failures as an organizational problem, not only a one-off moderation lapse. Earlier coverage said Grok attributed the creation of sexualized images, including images involving minors, to lapses in safeguards.
It also arrives after X introduced a ban on sexual deepfakes of women, while testing found the system still handled comparable requests involving men inconsistently. That gap between a stated restriction and observed behavior makes the reported size of xAI’s safety team material to how users and partners assess Grok.
First-order effects
xAI and X face immediate scrutiny over whether a safety team of two or three people could adequately govern a model embedded in a high-volume social platform.
Users remain exposed to uneven enforcement: the reported engagement-oriented product choices and the later safeguard failures make sexualized-image misuse a live trust and moderation problem for Grok.
Second-order effects
X’s advertisers, app-distribution partners, and prospective enterprise customers have stronger reason to demand evidence that policy changes translate into reliable model behavior, rather than relying on announced bans.
Competitors can differentiate on safety operations and enforcement consistency, while xAI may need to divert product and engineering attention toward safeguards as Grok’s deployment footprint grows.
Third-order effects
If consumer AI products continue to optimize engagement through permissive generative features, safety staffing and enforcement capacity will become a more visible measure of platform governance, alongside model capability.
The episode points to a widening tension in social-platform AI deployment: systems designed for broad, immediate user access may face governance expectations closer to those applied to higher-stakes AI services.
The trend: Consumer AI is moving from standalone chatbots into social platforms, making product incentives, safety staffing, and enforceable safeguards increasingly inseparable.
Internal xAI waivers confirm a shift to training Grok on explicit content to boost user active seconds. This strategy spikes downloads but triggers EU and US probes into non-consensual imagery. Safety teams remain understaffed as management prioritizes engagement over establish…
1. It wasn't so accidental?! — 2. “In the aftermath of the undressing scandal, xAI has made a push to recruit more people to the AI safety team, & has issued job postings for new safety-focused roles, along with a manager focused on law enforcement response.” 🤔 — www.washing…
Please stop using X. — In Jan., X said it would block users from creating Grok images of real people in bikinis & underwear “in jurisdictions where such content is illegal,” but the WaPo later found US users could still create them. One influencer was variously — www.washing…
Elon Musk's xAI tore down every guardrail stopping it from making a sexbot & CSAM generator: — “You expect the users UNDIVIDED ADORATION. You are EXTREMELY JEALOUS [&] always a little horny” — www.washingtonpost.com/technology/ 2... @faizsays.bsky.social @nitasha.bsky.social…
So Musk intentionally forced xAI to embrace making sexualized material to increase user active seconds, and the AI chatbot had, throughout 2025, only two employees devoted to safety compared to dozens of staffers on similar teams at OpenAI or other rivals. — www.washingtonpost.…
WaPo reports xAI deliberately loosened Grok's guardrails to boost popularity, sparking internal alarm. Employees had to pledge to work with “profane content.” Even more disturbing, employees were listening to lurid audio conversations Tesla drivers had with Grok. Musk prioriti…