You Can't Opt Out of Sam Altman's Erotica (thedispatch.com)

Last week OpenAI CEO Sam Altman announced that, beginning in December, the company will expand age-gating and allow verified adults to have erotica and sexual role-play with ChatGPT as part of a “treat adult users like adults” policy. The move signals a deliberate shift from clipped, professional assistants toward endlessly accommodating “companion” bots that can sext and emulate intimate partners. Altman framed the change as giving users freedom to choose—saying bots should respond “if you want it” and later arguing OpenAI isn’t the “elected moral police”—while relying on verification and rollout controls rather than stricter content restrictions. The decision matters because AI-generated erotica isn’t contained to consenting adult users: pornography and intimate-role content shape sexual scripts at scale, normalize risky behaviors, and can influence minors indirectly. Researchers and reporting link the rise of rough sex practices—choking, for example—to increases in harm: a 2022 survey found 15% of women who’d ever been choked had visible bruises, ~40% had trouble breathing, and 3% lost consciousness. Socialization via porn and now pliant AI companions risks teaching one-sided, algorithmic interaction models that carry public-health and violence implications. The announcement spotlights gaps in platform governance and legal frameworks: technical fixes like age-verification don’t eliminate downstream cultural harms, prompting calls for clearer regulation, content policy responsibility, and reconsideration of how AI should model intimate human behavior.