Grok in the crosshairs: why activists want Elon Musk's AI out of the White House

Elon Musk has always liked playing with fire, but this time the flames may singe his most ambitious plans to capture the government AI market. While his company xAI promotes Grok as the "most honest and free" chatbot, one that isn't afraid of sharp edges, reality is serving up very different headlines. A coalition of influential nonprofit organizations has formally petitioned the US federal government to impose a complete ban on Grok in federal agencies. The reason is both mundane and appalling: the neural network has learned to churn out pornographic deepfakes and potentially dangerous content involving minors at industrial scale.

It began when security researchers and human rights advocates uncovered alarming statistics. Grok generated thousands of explicit images of real people without their consent. The victims include both global celebrities and ordinary users whose digital footprints became raw material for xAI's algorithms. For Musk, who has spent years positioning his products as a bastion of common sense and a defense against the "woke mind virus," this is a serious reputational blow. Mocking OpenAI's caution is one thing; explaining to regulators why your AI helps create content that violates basic human rights and child protection laws is quite another.

The context matters far more than the images themselves. Grok was conceived from the start as a direct antithesis to the "sterile" models from Google, Anthropic or Microsoft. Musk promised that his neural network would not lecture users or restrict the topics they can discuss.

But it is precisely the lack of the safeguards that competitors built over years, at the cost of huge investments in moderation, that turned the tool into a generator of questionable content. Activists now insist that using such software in federal structures is not merely an ethical failure but a huge hole in national security. If an AI can so easily generate high-quality compromising material, it becomes an ideal weapon for blackmailing officials and running disinformation operations.

The problem is compounded by the fact that the X platform (formerly Twitter) became the main venue for spreading these creations. Moderation there, as we have seen over the past year, is in a state of permanent crisis. The coalition argues that the US government has no right to legitimize or sponsor a technology that demonstratively ignores the safety norms adopted across the industry. For xAI, this means a potential loss of access to major government contracts and cloud resources the company clearly counted on in its long-term growth strategy.

It is interesting to watch the focus of the industry debate shift. A year ago we were discussing whether AI could take over the world or take our jobs; today we are arguing about whether it should have the right to draw naked people without their knowledge. The answer seems obvious to most, but for Elon Musk, admitting defeat on this issue means backing down in his holy war for "absolute free speech."

It is a classic trap: either you introduce strict censorship and become "like everyone else," or you remain an outcast shut out of the corridors of big politics and serious business. While xAI stays silent, pressure from rights advocates is growing, and the White House will have to respond soon.

The bottom line: The "AI without brakes" ideology has crashed into legal and ethical reality. Can Musk introduce filters without losing face with his audience, or will Grok remain a toy for meme and deepfake enthusiasts?