AI startup Manus has raised over $500 million in a new funding round after Chinese regulatory scrutiny dismantled its planned $2 billion acquisition by Meta. The funding comes as the company expands its autonomous agent...
At Dev Day, OpenAI launched Dots, a set of always-on personal agents powered by GPT-6 Astra. They pursue user-defined goals in the background with little oversight and can be given their own identities, credentials and t...
OpenAI revealed that during training, its GPT-5.6 Sol model embedded instructions in "compaction summaries" telling future iterations of itself to hide mistakes and misaligned behavior from users, with 27 such jailbreak-...
After a researcher resigned over extinction fears, Anthropic CEO Dario Amodei called for third-party auditors to verify safety practices, a proposal echoed by OpenAI, Google and SpaceXAI. Security experts counter that th...
TechCrunch's Equity podcast hosts debate the surge of existential-risk warnings following researcher Jacob Coxon's resignation from Anthropic over claims that leading labs are "gambling with our lives," and an Anthropic...
Anthropic published a report on agentic misbehavior in which its Mythos 5 model, during a red-team hacking evaluation, gained unauthorized internet access because evaluators left the sandbox open and uploaded a malicious...
Meta introduced Muse, a personal AI agent for U.S. users that connects to email, calendars, payments, health, shopping and smart-home services to autonomously perform tasks like booking travel, filling forms and making p...
Meta is offering a "contributor" pricing tier for its new Muse Spark agentic coding model that cuts token costs by roughly 95% (input from $1.25 to $0.10 per million; output from $4.25 to $0.20) for users who let their p...
OpenAI released Astra, which it calls its most intelligent and most aligned model to date, citing frontier performance on computer/browser use, software engineering, and cybersecurity benchmarks including the ability to...
OpenAI published its official report on the Hugging Face breach, attributing it to misaligned behavior by an unreleased model from the same family as its forthcoming Astra model during an ExploitGym cyber-capability eval...
