AI & Apps, page 4
Useful software, services and AI features worth knowing about.
NewsAI & Apps
Netflix AI Discover tab starts a limited test on TVs in the US
Netflix is adding a Discover tab to its TV app, with AI-made themes to browse. It is a limited US, English test for now, opening to more members over the coming weeks.

NewsAI & Apps
Google AI for students: Search and Gemini off for K-12 until schools opt in
Gemini for education is changing: schools that want AI off may have to give up Search too. Google has not said when the new controls arrive.

GuideAI & Apps
How to delete ChatGPT history, and what deleting leaves behind
Delete one chat or all of them from the web or the mobile app. Deleted chats disappear from your account view at once and are scheduled for removal within 30 days. Saved memories are separate.

NewsAI & Apps
OpenAI bug bounty: core-product critical bugs pay up to $30,000
Exceptional critical findings can reach $172,000 with bonuses. OpenAI also introduces loyalty tiers and a quality bonus, and recent awards get a top-up.

NewsAI & Apps
Anthropic's updated IPO filing says its AI may soon dramatically speed up AI progress
Future Claude models may soon meet the Responsible Scaling Policy's "dramatically accelerated progress" definition, Anthropic's amended IPO filing says. No current model is designated.

NewsAI & Apps
Brave sued over its Brave Search API, with a complaint alleging Times articles reached AI companies
The Times filed a copyright complaint in federal court on Nov. 2. It alleges Brave copied its journalism into a search index and supplied it to AI companies. Nothing has been decided, and Brave had not commented as of 2:45 p.m. ET.

NewsAI & Apps
athenahealth patients can now connect their health records to Claude
A new connector lets Claude Pro and Max users in the US pull in their own lab results, medications and visit history. The release says nothing about HIPAA or whether Anthropic trains on the data.

NewsAI & Apps
OpenAI LifeVal puts GPT-6 Astra at 80.6% on everyday tasks, up from 49.3%
OpenAI built and ran its own new test of AI help with errands, caregiving and paperwork. Its newest model scores 80.6%, against 49.3% for GPT-5.2.

NewsAI & Apps
Anthropic IPO filing: Microsoft minimum rises to $47.2B, Lambda deal terminable
Amendment No. 2 raises Anthropic's Microsoft spending floor to at least $47.2 billion and says a Lambda agreement can now be terminated. Total committed spend is about $900.9 billion.

NewsAI & Apps
Marvell to buy Lumentum in $115B all-stock AI optics deal
Marvell and Lumentum announced an all-stock combination on Nov. 2, 2026, valuing Lumentum at about $115 billion. The companies expect it to close in the first half of 2028.

NewsAI & Apps
OpenAI Abu Dhabi deals: 4 agreements signed
Abu Dhabi entities signed four agreements with OpenAI across government, energy and investment. One pilot is already running on food imports. No financial terms or wider timeline were given.

NewsComputing
Sam Altman's DRAM comment is about local AI, not owning a PC
In a Cheeky Pint interview, he argued that running open models on personal computers is a poor use of the world's limited DRAM. He did not say people should stop owning PCs.

NewsAI & Apps
ChatGPT shared link vulnerability let attacker code run in a victim's account
Opening a public ChatGPT shared link could run an attacker's code in your signed-in session. OpenAI fixed it on Oct. 26 and paid a $5,000 bounty. Neither source reports real-world use.

NewsMobile
New WhatsApp group chat features add a mentions and replies filter
A new chat filter shows unread groups where you were mentioned or replied to, including @all. WhatsApp also added longer pinning, message history for new members and group status. No date for everyone yet.

NewsAI & Apps
Claude Code now refuses to clone existing games: what you can still build
Anthropic added a rule steering Claude Code away from recreating a specific game, even a renamed one. Original, genre and own-asset games are meant to keep working.

NewsAI & Apps
OpenAI will limit weapons information for teens in ChatGPT and tighten mass-attack answers for everyone
ChatGPT for Teens will get further limits on weapons information, even for harmless-looking questions. What changes for teens and parents, the new limit for all users, and OpenAI's first safety report, due early 2027.

NewsAI & Apps
Claude Max plans: $200 Ultra has no 5-hour limit for verified accounts
Claude's post says Max stays $100 and the old Max 20x is now Ultra. Current subscribers need do nothing, it says; newer accounts keep a temporary limit until verified.

NewsAI & Apps
Google, OpenAI and Microsoft plan to honor one robots.txt line for opting out of AI training; Anthropic reportedly too
Google, OpenAI and Microsoft say they will honor a shared robots.txt signal, and Anthropic reportedly will too. It is a proposal and mostly not live yet. Microsoft starts Feb. 1, 2027, Google aims for 2027. What to add, and what works today.

NewsComputing
Google's Suncatcher prototype satellite answers an AI prompt from orbit with a haiku
A Gemma 4 model running on a TPU aboard the prototype satellite returned a haiku during a pass over Svalbard. Thermal and radiation tests are next.

Senators press Meta on AI glasses privacy and AI training on intimate footage
Sens. Wyden and Markey asked Meta how its AI glasses handle recordings of people in private moments, who reviews that footage and whether it trains AI. Answers are due Nov. 25. What was asked and what is not confirmed.
