Judge Bans ChatGPT User Without First Amendment Review, Raising Free Speech Concerns

California court order suspending user's access to AI platform skips constitutional analysis, setting dangerous precedent

Apr. 15, 2026 at 3:51pm

Got story updates? Submit your updates here. ›

A highly detailed, 3D illustration of a glowing, neon-lit ChatGPT icon or logo, surrounded by a swirling array of digital data streams, circuit boards, and other cybernetic elements, conveying the power and complexity of AI-powered speech platforms.As courts grapple with the constitutional implications of restricting access to AI-powered speech platforms, the legal and ethical complexities of this emerging technology are brought into sharp focus.Los Angeles Today

A California judge granted a temporary restraining order requiring OpenAI to suspend a user's access to ChatGPT, based on allegations the AI tool enabled his stalking campaign against an ex-girlfriend. However, the court did not conduct a First Amendment analysis, despite OpenAI citing Supreme Court precedent on the constitutional right to access online platforms. Legal experts warn this sets a concerning precedent where AI platform access could be restricted without proper consideration of free speech protections.

Why it matters

This case highlights the legal gray area around regulating AI-powered speech and the potential for courts to infringe on constitutional rights in the name of public safety. As AI becomes more integrated into everyday communication, there are growing concerns about how to balance platform moderation with fundamental free speech principles.

The details

According to the report, San Francisco Superior Court Judge Harold Kahn granted a temporary restraining order on April 13 requiring OpenAI to suspend a user identified as John Roe from accessing ChatGPT. This was based on allegations that Roe used the AI tool to enable a stalking campaign against an ex-girlfriend, known as Jane Doe. However, the court apparently did not consider OpenAI's arguments that suspending Roe's access would violate his First Amendment rights, citing the Supreme Court's Packingham v. North Carolina ruling that the internet is the 'modern public square' where free speech must be protected.

  • On April 13, 2026, Judge Harold Kahn granted a temporary restraining order requiring OpenAI to suspend John Roe's access to ChatGPT.
  • A preliminary hearing in the case is scheduled for May 6, 2026.

The players

John Roe

A user of the ChatGPT AI platform who was banned from accessing the service by a California court order, based on allegations he used the tool to enable a stalking campaign against an ex-girlfriend.

Jane Doe

The ex-girlfriend of John Roe, who filed a lawsuit alleging Roe used ChatGPT to generate fabricated psychological reports about her and distribute them to her family and colleagues.

Judge Harold Kahn

A judge in the San Francisco Superior Court who granted a temporary restraining order requiring OpenAI to suspend John Roe's access to ChatGPT.

OpenAI

The company that operates the ChatGPT AI platform, which was ordered by the court to suspend a user's access to the service.

Eugene Volokh

A First Amendment scholar who observed the court hearing and noted that there was no meaningful discussion of the user's speech rights by the court.

Got photos? Submit your photos here. ›

What they’re saying

“there was no meaningful discussion of the user's speech rights by the court.”

— Eugene Volokh, First Amendment scholar

What’s next

The May 6 preliminary hearing will determine whether this court order becomes extended precedent, though the case may also transfer to California's coordinated ChatGPT litigation proceedings.

The takeaway

This case highlights the legal complexities around regulating AI-powered speech and the potential for courts to infringe on constitutional free speech protections in the name of public safety. As AI becomes more integrated into everyday communication, there are growing concerns about how to balance platform moderation with fundamental free speech principles.