OpenAI Faces Internal Turbulence as ChatGPT ‘Adult Mode’ Delayed and Hardware Chief Resigns

OpenAI has postponed the launch of ChatGPT’s “adult mode,” a feature intended to allow verified adult users to access erotica and other age-restricted content through the chatbot. The capability was originally announced by CEO Sam Altman as part of the company’s effort to introduce stronger age-verification controls and expand content options for adult users.

The feature had already been delayed from its initial rollout and has now been pushed back again as the company prioritizes improvements to ChatGPT’s core capabilities. OpenAI said development efforts are currently focused on enhancing the model’s intelligence, personality, and proactive behavior. The company indicated that its principle of treating adult users differently from minors remains part of its long-term approach but said more time is required to ensure the feature is implemented responsibly.

The delay comes amid internal and public scrutiny following OpenAI’s recent agreement with the U.S. Department of Defense, which allows the company’s AI technology to operate in classified environments. In response to the deal, Caitlin Kalinowski, who led OpenAI’s hardware team and previously worked on augmented reality hardware at Meta, announced her resignation.

OpenAI confirmed Kalinowski’s departure and said its defense agreement establishes safeguards intended to prevent domestic surveillance and autonomous weapons use while enabling responsible national security applications of artificial intelligence.