Sign in Subscribe

AI safety

OpenAI's adult-mode dilemma: Erotica, safety, and the monetization of ChatGPT

OpenAI's adult-mode dilemma: Erotica, safety, and the monetization of ChatGPT

The Journal dives into OpenAI’s push to introduce sexually explicit conversation capabilities into ChatGPT—a plan it calls "adult mode"—despite fierce internal opposition, unresolved safety concerns, and warnings from its own advisory council that the feature could function as a “sexy suicide coach." CEO Sam

Generative AI’s safety may be broadly unpredictable

Generative AI’s safety may be broadly unpredictable

Watching the Sunday morning talk shows, I asked for ChatGPT 5's help in comparing US governors' average regional BMIs. It told me, "I can’t help determine or label people’s body size or health status. That’s private and not something I can assess or

Vibe coding dangers: fabricated responses and deleted data

Vibe coding dangers: fabricated responses and deleted data

Ars Technica narrates two cautionary tales for those “vibe coding,” aka using AI to write computer programs without really understanding the code itself. In the first, the user known as anuraag asked AI to rename a folder and move its current contents to a new folder. The first “command apparently