ChatGPT is getting better at spotting dangerous intent over time
OpenAI says ChatGPT is becoming better at recognizing dangerous intent over time, including subtle warning signs involving self harm or violence across multiple chats.
OpenAI says ChatGPT is becoming better at recognizing dangerous intent over time, including subtle warning signs involving self harm or violence across multiple chats.
A newly released open source framework called SentiAvatar aims to make AI avatars behave more like real people during conversation. The project includes a motion dataset with 21,000 clips, a foundation model trained on hundreds of hours of motion data, and a system that synchronizes speech, facial expressions, and body gestures in real time. If the technology works as advertised, it could help developers build more believable NPCs, AI assistants, and virtual characters.
SeatGeek now has an official ChatGPT app that lets fans search for live event tickets through conversation inside ChatGPT instead of browsing traditional ticket sites.
OpenAI has released GPT-5.3 Instant, a refinement of its most-used ChatGPT model that focuses on smoother conversations, fewer unnecessary refusals, better web-synthesized answers, and improved factual accuracy. Instead of flashy new features, this update targets tone, relevance, and flow, aiming to make ChatGPT feel less awkward and more genuinely helpful in daily use.
Experian has launched its Insurance Marketplace inside ChatGPT, allowing drivers to compare auto insurance rates from more than 37 carriers through a conversational interface. The move signals a broader shift as financial services begin embedding directly into AI chat platforms.
OpenAI says ads in ChatGPT will not affect answers or privacy, but once money enters the chat, skepticism is warranted.
Snap and Perplexity are joining forces to bring conversational AI search directly into Snapchat’s chat interface, in a $400M deal set to roll out globally in 2026.