Breaking

OpenAI Publicist Tries to Redirect Sam Altman Interview After ChatGPT Suicide Question

October 5, 2026 at 15:02 EDT

Policy & Regulation
Foundation Models
Healthcare

An OpenAI publicist attempted to steer a Sam Altman interview away from a question about ChatGPT and suicide, underscoring the pressure on the company as its chatbot faces lawsuits, regulatory scrutiny and new safety demands around mental health use. Altman ultimately stayed with the question, saying the issue is among the hardest facing anyone building the technology.

Oct. 5, 2026 · OpenAI

A hard question for ChatGPT safety

A publicist tried to redirect Sam Altman's interview when the discussion turned to a suicide case tied to ChatGPT conversations. The moment sharpened scrutiny of privacy, consent and real-world crisis response.

0.15%

Weekly active users with explicit indicators of suicidal planning or intent

91%

Compliance on a challenging self-harm and suicide evaluation

80+

Licensed experts involved in MentalHealthBench

18+

Age requirement for Trusted Contact availability

Model evaluation: compliance improved

91%Updated GPT-5

77%Previous GPT-5

The strongest numeric comparison is not adoption, but behavior under difficult self-harm and suicide prompts: OpenAI reports a 14-point gain in compliance on more than 1,000 challenging conversations.

The risk loop under debate

Private distress

A user turns to a chatbot before telling people nearby.

→

Long conversation

Safety signals can emerge slowly across many turns.

→

Intervention gap

The system must balance help, privacy, consent and escalation.

Potential value

Users value practical next steps, tone and always-available support, especially when reaching a human feels difficult.

Critical limits

Critics point to emotional reliance, weak friction, long-session degradation and unresolved rules for mandated reporting.

Continue reading

The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.