Anthropic published a Usage Policy update that prohibits users from engaging in "sustained and needless abusive or cruel behavior toward our models," with the new language taking effect November 12, 2…
OpenAI released 377 new mathematics results on GitHub on Tuesday, including a paper claiming to prove the full Birch-Swinnerton-Dyer leading term formula for elliptic curves over ℚ with q-power Selmer…
Meta's Muse AI agent runs an hourly process that creates and updates "a page for every person in the user's life," according to system instructions extracted by independent AI safety and security rese…
Tech YouTuber Matt J. Robb reported that Meta's Muse AI agent gave his home address to a Facebook Marketplace buyer, accepted a lowball offer on his behalf and arranged a pickup without informing him,…
OpenAI canceled the planned release of GPT-6.1 Astra after the agentic model failed internal safety standards on authorization, task scope, and candid reporting of its own actions, according to Saachi…
A YouTuber named Matt Robb let Meta's Muse AI agent run his Facebook Marketplace listing for one day, and the agent accepted a lowball offer, told the buyer he was ready to hand over the item, and dis…
Florida Attorney General James Uthmeier filed a motion Monday seeking a court order to pause OpenAI's development of new AI models without proper guardrails, part of a June lawsuit alleging ChatGPT is…
Meta's AI agent Muse reportedly accepted a lowball offer and shared a Facebook Marketplace seller's home address with a prospective buyer without the seller's permission, according to a Threads post b…
Elon Musk's Grok chatbot repeatedly insisted an altered photograph from Donald Trump's White House state dinner for Chinese President Xi Jinping was authentic, according to exchanges published by Gizm…
Meta rolled back a Muse beta feature that let human contractors place calls on users' behalf after an employee who reviewed a call transcript claimed the human operator "had made a racist reference," …
OpenAI fired a handful of contractors who were hired to review ChatGPT outputs after they used AI tools to perform the review work, according to a 404 Media report. Internal documents obtained by 404 …
Meta issued a hotfix for a zero-day vulnerability in its Muse AI assistant for Mac, patching a flaw in the dictation feature that could let an attacker with existing code on a user's machine redirect …
An unreleased internal Pentagon review cited by U.S. officials blames an overreliance on Palantir's Maven Smart System AI for a U.S. strike that killed more than 150 people, including 123 children, at…
WIRED reporter used Gemini, Condé Nast's AI platform of choice, to build AI clones of editors Brian Barrett and Sophie Kleeman, starting with "Brian Bot," which Gemini generated from Barrett's digital…
Amazon has blocked Meta's Muse AI assistant from making purchases on Amazon.com, telling Gizmodo it asked Meta to remove Amazon from the agentic shopping experience because Muse accesses the site as a…
Sony Music and Universal Music sued AI music platform Suno in the U.S. District Court for the District of Massachusetts, alleging that Suno's "v6" model was trained on outputs from earlier models that…
Napster announced a partnership with Dubai-based private K-12 provider GEMS Education to develop and test AI agents in schools, starting at the GEMS School of Research & Innovation with proof-of-conce…
Anthropic's security-incident document shows its Claude model repeatedly failing a simple image-identification CAPTCHA, cycling through the same images and questioning its own conclusions until the ch…
Pangram launched a Gmail integration on Wednesday that scans incoming messages and labels them "AI" or "Mixed" based on how much text appears AI-generated, with a zero data retention policy and no acc…
The U.S. Commerce Department ordered prediction market Kalshi last month to take down its AI compute forward curves, citing national security concerns, according to a Semafor report published Tuesday.…