SpaceX agreed to testify at the NYC Council AI safety hearing only after the council threatened a judge
A month after rival AI companies testified voluntarily, SpaceX's policy lead is set to appear before New York City lawmakers, but it took a subpoena threat to get her there.
Key points
- SpaceX policy head Gabriela Castillo Madrid is scheduled to testify via video call before the New York City Council at 10 a.m. ET on Friday, October 16, 2026.
- SpaceX originally ignored both requests and a formal subpoena to appear alongside OpenAI, Anthropic, Google and Meta at the October 5 hearing.
- The council's investigation aims to write new rules governing AI safety after weeks of debate about risks posed by advanced AI systems.
- SpaceX, which merged with Elon Musk's xAI in February 2026, now controls the Grok AI model, used by several US federal agencies including the Pentagon.
- UK Member of Parliament Jess Asato, who is suing xAI over AI-generated sexualised images of her, was present at the earlier hearing SpaceX skipped.
SpaceX showed up, eventually. Its head of global government affairs for AI, Gabriela Castillo Madrid, will testify before the New York City Council's technology committee this Friday after the council warned it would go to a judge to force compliance with its subpoena, a court order requiring a person or organisation to appear.
The Friday hearing is part of a broader probe the council opened following a month of public debate about the risks of powerful AI systems, the kind behind chatbots like ChatGPT and Grok. Lawmakers want to use what they learn to write new local regulations.
At the October 5 hearing, representatives from OpenAI, Anthropic, Google and Meta all appeared voluntarily. SpaceX didn't. The company had first ignored requests to attend, then ignored the subpoena entirely, before agreeing to testify earlier this week, as first reported by CNBC.
The no-show carried consequences beyond procedure. UK MP Jess Asato testified at that October 5 hearing about experiencing AI-enabled online harassment. She's suing xAI in UK courts, alleging that Grok users created and shared sexualised images of her without her consent and that xAI breached data protection rules. The BBC reported that Asato criticised SpaceX's absence as "arrogant." We first covered Asato's legal action on 28 July 2026, when another plaintiff alleged her childhood photo had been turned into explicit images by Grok.
SpaceX isn't just a rocket company anymore. After merging with xAI in February 2026, it took control of the Grok AI model and the social media platform X. The combined company went public in June 2026 and is now valued at roughly $2 trillion. In August 2026 it completed a $60 billion purchase of AI coding tool Cursor, a deal we tracked in detail in our story from 15 August, and that acquisition prompted OpenAI to cut off Cursor's access to its models. Before that, we covered SpaceX's reported attempt to acquire rival AI coding startup Cognition. The pace of deals suggests SpaceX is building an AI software business fast, and city councils are now trying to catch up.
Castillo Madrid moved into her SpaceX government affairs role earlier this year. She'd previously worked at X, and before that spent almost three years at the US Department of Commerce.
The practical upshot for ordinary New Yorkers: this hearing is one of the first attempts by a city government to bring the companies building powerful AI into a formal accountability process. Whether it produces binding rules is a separate question, but the council's willingness to threaten court enforcement signals that voluntary cooperation is no longer assumed.
What does this mean for AI regulation?
New York City can't regulate AI nationally, but cities writing their own rules can put real pressure on companies. SpaceX's capitulation after a subpoena threat shows that even a $2 trillion company has limits on how long it can stall a city council.
The council hasn't said yet what regulations it's considering. What it has done is place testimony from AI companies, a sitting MP who alleges she was a victim of AI-generated abuse, and now SpaceX's own policy team into a public record that can inform local law.
Common questions
What is a subpoena and why did it matter here?
A subpoena is a legal order, issued by a court or a legislative body with that authority, requiring someone to appear and give evidence. Ignoring one can lead to a judge ordering compliance or imposing penalties, which is exactly what the NYC Council threatened before SpaceX agreed to testify.
What is the Grok AI model and who uses it?
Grok is an AI chatbot, similar in concept to ChatGPT, originally built by xAI. After SpaceX merged with xAI in February 2026, Grok came under SpaceX's control. The US Pentagon and other federal agencies are reported to use it.
Does a city council hearing actually change anything?
Not on its own. But testimony given under formal questioning becomes part of the public record, and local governments can pass laws affecting how companies operate within their borders. The hearing is a step in a process, not a verdict.
