Google is simplifying the Gemini app's thinking level picker by removing the explicit 'Standard' option and showing only 'Extended thinking' for a one-step process, with the change rolling out widely on Android and the w…
A developer launched x402, a pay-per-call API platform offering 337 utility tools including DNS lookups, QR codes, AI chat, and code review, with a free trial of 3 calls and paid plans starting at €0.50 for 100 calls or …
South Korea's benchmark Kospi index fell as much as 7.6% to its lowest since April 20, led by a 11% plunge in SK Hynix Inc. and a more than 9% drop in Samsung Electronics Co., as deepening AI fatigue and concerns over Nv…
A user identified as Bitu79 on Google's AI Studio forum claims that the 'Delete' button in Google AI Studio does not remove chat data from Google's servers, only deleting the JSON file from Google Drive while the chat se…
A user identified as Bitu79 reports that Google AI Studio's delete dialog claims prompts are permanently deleted after 30 days or when the trash is emptied, but testing shows that after emptying the Drive trash and using…
A user identified as bitu79 claims Google illegally retains customer data in Google AI Studio, alleging that pressing the delete button only moves a JSON file to Google Drive's trash rather than deleting backend data. Th…
Anthropic CEO Dario Amodei published a formal position on July 27, 2026, stating he does not advocate for a ban on open-weight AI models but warning that Chinese frontier models like Moonshot AI's Kimi K3, released for f…
Nvidia announced the Open Secure AI Alliance on Monday, a coalition with Palantir, Microsoft, OpenClaw, Salesforce, and SpaceX to develop open technologies for AI security, while also entering a long-term partnership wit…
Nvidia and Safe Superintelligence Inc. (SSI) announced a long-term partnership on July 27 that will give Ilya Sutskever's lab access to Vera Rubin systems and increase its compute capacity tenfold. Nvidia disclosed an in…
A viral post by user ali@waterloo_intern22580 on X, garnering 872.3K views, traces the evolution of Transformer models from GPT-2 to Kimi3, earning praise from Dibyadeep Saha as one of the best write-ups on the topic.
Nvidia launched the Open Secure AI Alliance, a global technology alliance of around 40 companies including Microsoft, Dell, CrowdStrike, SpaceX, and Hugging Face, to identify vulnerabilities and strengthen AI security. N…
LM Studio Bionic, the agentic workflow platform for open models, now supports Moonshot AI's Kimi K3, a 2.8-trillion-parameter model with a 1-million-context window. The addition gives users access to what many consider t…
Anthropic's Claude Fable 5 and OpenAI's GPT-5.5 Pro each generated verified solutions to an unsolved number theory problem involving the absolute Galois group of the 2-adic numbers, marking the second solved problem on E…
Google co-founder Sergey Brin revealed that Gemini, Google's own AI model, was on an internal list of coding tools that engineers were banned from using, and he had to enlist CEO Sundar Pichai to get the restriction lift…
Anthropic CEO Dario Amodei stated Monday that his company has never advocated for a ban on open-weight AI models, pushing back against industry speculation, but he expressed fears that Chinese AI could enable biological …
PromptTrace, a free hands-on lab platform for learning LLM security through prompt injection, AI red teaming, RAG poisoning, and tool exploitation, has launched with 10 labs, 17 CTF levels, and 9 learning modules. The pl…
Google shipped four production-readiness upgrades to Gemini API Managed Agents on July 7, including background execution that allows agents to run asynchronously server-side without a persistent HTTP connection, and remo…
A developer built Start Story, a reusable AI workflow that automates Azure DevOps work items from acceptance criteria to pull request creation while keeping developers in control. The system separates deterministic scrip…
OpenAI, Google, and Anthropic are absent from the Open Secure AI Alliance, a new industry group led by Nvidia aimed at improving AI security. The alliance includes over 20 founding members such as AMD, Intel, Microsoft, …
Asking a large language model to generate a confidence score for its own response is completely useless, argues an unnamed author who has observed the pattern across multiple companies. The author contends that there is …