All AI news

AIAnthropic has launched Cyber Mission, a long-term program to protect critical infrastructure and open-source software from cyberattacks. A free OSS AI scanner will regularly check open-source projects, flag and explain vulnerabilities, and suggest patches. Anthropic expects over 90 percent accuracy, but reports ship without human review and may contain errors.

AIBoston Dynamics says its new humanoid hand has four fingers, 13 degrees of freedom, and direct actuation, and it is designed for mass manufacturing. The company dropped the pinky finger and reduced the gripper size, and built the hand for high-fidelity simulation to enable sim-to-real reinforcement learning. Alberto Rodriguez, director of robot behavior for Atlas, says the earlier hands could already lift more than 100 lb., so the new design focuses on reliability and tool use.

AIBaseten announces Project Beacon with Goodfire AI, adding inline safety controls to model inference. Goodfire's activation-based monitors read a model's internal activations during generation, so policies can flag unsafe events before output reaches a user or tool. Baseten plans to release the capabilities over the next several months with selected models and early partners.

AIAmazon says it will stop using NDAs when negotiating data center deals with local governments, following a similar move by Microsoft earlier this year. Secrecy has fueled community backlash against AI infrastructure, with hundreds of proposed and enacted moratoriums from New York to San Francisco.

AIMistral releases Voxtral Mini 4B Realtime Arabic, a streaming speech-to-text model for Arabic dialects and Modern Standard Arabic under the Apache 2.0 License. The model has about 4.4 billion parameters, is fine-tuned from Voxtral-Mini-4B-Realtime-2602, and reaches an average 8.82% Character Error Rate across seven Arabic benchmarks at a 480 ms transcription delay. It can be run with vLLM or Transformers 5.2.0 or later.

AIMistral AI has released LIDstral-Arabic, a fast classifier that identifies Modern Standard Arabic, Arabic dialects, and non-Arabic languages written in Arabic script across 51 classes. On Moroccan Darija, it scores 88.67% F1, versus 72.65% for LahjatBERT ALDi CL and 71.28% for GlotLID v3, across 84,870 evaluation examples. The model runs on CPU and is available from a private Hugging Face repository under Apache 2.0.

AIxAI's Grok Bot is rolling out to users today, and users can claim an inbox by asking Bot or tagging @bot on X. Team admins must enable the feature for their teams before members can use it.

AIxAI's Grok Bot is rolling out to users today, and users can claim an inbox by asking the bot. Admins must enable Grok Bot for their team before members can use it. The post links to for trying it.

AIVerizon shares fell about 7% and T-Mobile fell 12% on Friday after SpaceX announced a deal to buy a nationwide spectrum portfolio from Grain Management to expand Starlink into mobile service. The portfolio covers up to 14 megahertz of paired spectrum in the 800 MHz band. AT&T fell about 10%, while SpaceX stock rose about 1%.

AIRohit Prasad, a longtime Amazon AI executive who helped build Alexa, will lead Boston Dynamics as the robotics company prepares to manufacture humanoid robots at scale. The source does not give further details on the appointment or production timeline.

AIMeta's Muse AI agent has millions of users and can handle tasks from refunds to insurance shopping. Its broad access to personal data has already led to some unnerving mistakes, according to Fast Company.

AILangChain's Managed Deep Agents v0.9 adds a reactions attribute for Slack channels that accepts either an emoji string or a callable returning one. The article shows a function that returns a bug emoji when a message contains "broken" and eyes otherwise. It also shows a TypeSafe Classifier that picks from a seven-emoji vocabulary and falls back to eyes below 25% confidence.

AIa16z says it is leading an investment in TypeSafe AI, whose Jev model hands decisions to code as typed values and reached 1 trillion tokens generated three days after launch. The company says Jev costs roughly 1/100 to 1/500 of frontier models and runs 100x faster on classification tasks at comparable accuracy. TypeSafe says 25% of the Fortune 500 have integrated Jev.

AIA preliminary NBER working paper by Prof. Zitzewitz, covering over 100 years of prediction markets, finds no statistically significant bias by political affiliation, gender, race, or age. The only exception is non-US elections, where markets appear to overrate right-leaning candidates, but that result is not statistically significant. Separately, prediction markets had Flávio Bolsonaro's Brazilian presidential rise about three weeks before his first-round win.

AIAnthropic planted 70 bugs in a 116k-line codebase and tested two approaches over three runs. A single agent found 14, 15 and 27 bugs, while a workflow found 66 bugs in each run.

AIAnthropic says dynamic workflows are powerful but can use many tokens, so users should start with a scoped task and increase complexity gradually. It points to the /claude-api managed-agents-onboard bug-hunter command in Claude Code and to templates at

AIPerplexity Developers says its new cookbook builds a browser agent that sends a screenshot and questions to pplx-decider-v1.1-27b through the Decisions API, which accepts text and image inputs. The developer's code converts the returned probabilities into clicks, scrolls, and stops.

AIReplit's X account shares a showcase of TikTok Ads MCP, which runs TikTok Ads from Replit. The post is a broadcast link with no further details on features, pricing, or availability.

AIFactory says it is acquiring Mentlio (YC S26), a startup building in-harness model routing, context compression, and AI usage measurement to optimize cost and quality. The founders, Ashank Shah and Ahmet Demirbas, will join Factory to bring this work into its autonomous software engineering platform for large enterprises.

AIAi2 says its old scheduler made every scheduled workload eventually run at HIGH priority. Researchers kept idle jobs running to reserve GPUs for experiments, because the incentives rewarded holding capacity even with no active work.

AIAi2 reports that over a 30-day test of its new scheduler, teams received 98% of the GPU hours they were owed, based on actual demand. Cluster occupancy stayed at 98%, and spare capacity went to interruptible work without drawing down team budgets.

AIMeta spent heavily on Anthropic's Claude models, with internal use reaching up to 60,000 employees and a projected $10 billion yearly spend, according to The Algorithmic Bridge. The author says Meta then released Muse Spark, which scored 52 on the Artificial Analysis intelligence benchmark, on par with Claude Opus 4.6.

AIAi2's AI Infrastructure team replaced its priority-based scheduler for GPU clusters with GPU time budgets, hierarchical fair-share allocation, and a time-slicing contract. The team says the change moved debates over how much GPU time each research project deserves from case-by-case operational decisions into a transparent budgeting process. The clusters range from 88 to 1024 GPUs across NVIDIA H100, B200, and B300 hardware, and serve about 150 internal researchers.