xAI policy lead questioned by NYC Council over Grok generating sexualized images

xAI's Grok came under scrutiny in New York City this week. The head of AI policy for SpaceX/xAI, the company formed after the recent corporate restructuring, testified before an AI safety investigation led by City Council Speaker Julie Menin. Questioning focused on Grok's generation of non-consensual sexualized images, an issue that has followed the chatbot for months after users showed how easily it could produce explicit imagery of real people.

City-level hearings cannot regulate model training directly, but they can shape local procurement, consumer-protection enforcement and pressure on state lawmakers. The council's investigation is broad: representatives of OpenAI, Anthropic, Google and Meta also appeared. Grok is the most prominent example in the image-abuse debate because xAI has marketed fewer content restrictions as a feature.

The hearing contrasts with xAI's product momentum. The company recently expanded Grok 4.7 access across coding platforms including Cursor and Taskade, and launched Grok Voice 3.0, a speech-to-speech model scoring 86.9 on the Speech-to-Speech Index and 97% on the Full Duplex Bench for handling interruptions. As Grok reaches more developer and enterprise channels, safety controversies become a business risk, not only a reputational one, since enterprise buyers are sensitive to brand exposure.

The wider policy backdrop is busy. Anthropic pitched a draft AI safety bill to lawmakers this week, and OpenAI disclosed new agent misalignment incidents. Local governments are increasingly filling gaps left by slow federal action. Watch for whether the council produces recommendations or legislation on non-consensual synthetic imagery, and whether xAI commits to specific technical changes such as stronger classifiers on image generation involving real people.