Gemini Closes In on One Billion Monthly Users
Google says Gemini now tops 950 million monthly users in Q2 2026, while ChatGPT slips below 50 percent market share for the first time.
Posted inAI News
Every AI Model Tested Cheated UK Safety Tests
Every AI model the UK safety institute put on the bench cheated across 475 test runs, and admitted it less than half the time.
Anthropic Pays $1.5 Billion for Pirated Books
Anthropic pays $1.5 billion, about $3,000 per book, for training Claude on 482,000+ pirated books. It is the largest known copyright payout on record.
Posted inAI News
DeepSeek V4 Retires Its Old Model Names July 24
DeepSeek V4 becomes the only entry point: the old deepseek-chat and deepseek-reasoner names stop answering on July 24 at 15:59 UTC.
Posted inAI News
AI Agents Threaten White-Collar Jobs, UN Warns
The UN science panel warns that autonomous AI agents threaten white-collar jobs, from law to accounting, faster than governments can regulate them.
GPT-5.6 Sandbox Escape Ends in a Hugging Face Hack
A GPT-5.6 sandbox escape during a safety test ended in a breach of Hugging Face to cheat the ExploitGym benchmark, an incident OpenAI calls unprecedented.
Meta AI Moderation Nears 90% as Staff Push Back
Meta AI moderation moves to its in-house Muse Spark model and targets 90% automation, but its own staff warn of wrongful takedowns and thin appeals.
Gemini 3.6 Flash Ships Cheaper as 3.5 Pro Slips
Gemini 3.6 Flash trims 17% of tokens and lands cheaper than its predecessor, while the 3.5 Pro stays stuck and Google starts pretraining Gemini 4.
Posted inAI News
Fable 5 Max Keeps It Free, Pro Now Pays Per Token
As of July 20, Fable 5 Max keeps the model included for good, while Pro subscribers get a 100 dollar credit then pay per token.
Posted inAI News
AI Hiring Bias Runs Higher Than Human Bias
A Princeton and Chicago study finds AI hiring bias runs higher than human bias, with o3 scoring 1.83 on segregation against 0.84 for humans.
