AI.info

The Pulse — Page 56

Browse The Pulse on AI.info.

• DistroKid Confirms UMG-Linked Takedowns as Artists Report Lost Songs

  DistroKid told The Verge it removed a small number of recordings in response to claims from Universal Music Group. Musicians say original tracks and licensed releases have disappeared without clear notice or explanations.

• Researchers Map a Decade-Long Pause for Frontier AI Training

  A working group’s report examines how controls on training chips could support a long pause in frontier AI training. Its feasibility assessment depends on governments cooperating on the proposed governance and verification measures.

• Agility Begins Nasdaq Trading With More Than $300 Million in Conditional Digit 5 Orders

  Agility began trading on Nasdaq after completing its business combination with Churchill Capital Corp XI, raising approximately $465 million in gross capital. Agility has more than $300 million in committed multi-year orders for Digit 5, subject to contractual milestones, product features and specifications.

• Thomson Reuters Faces Antitrust Suit Over Casetext Shutdown

  The Rubin Law Firm alleges that Thomson Reuters bought Casetext and dissolved the legal research service, leaving the firm with higher monthly rates. Thomson Reuters completed its $650 million acquisition in August 2023, saying the deal supported its generative AI strategy.

• Atomic Raises $12.5M to Automate Supply-Chain Decisions

  Atomic raised $12.5 million in a Series A led by Klass Capital and Madrona. The former Tesla team says its software already automates 90% of purchasing for DoorDash’s DashMart business.

• DOE Selects Four Lab Robotics Projects in a $30 Million Program

  The Department of Energy selected four national laboratory-led robotics projects on October 8, 2026, under a program with $30 million in planned funding. The projects target reusable lab-robot skills, reliability testing, connected experiments and autonomous operations at photon-science facilities.

• NSF Sets Out New AI Science Programs, but Details Are Still Pending

  The National Science Foundation announced new programs on October 8, 2026, including research competitions and support for digital-first scientific instruments. More details will follow, and NSF projects up to $75 million for the instrument effort over five years.

• Chinese Resellers Route Claude Into a Gray Market

  A gray market of Chinese resellers is selling access to Anthropic’s Claude through pooled accounts and overseas servers. The Information’s October 2 report describes how sellers acquire accounts, evade bans and expose some customers’ data to intermediaries.

• DOJ Launches Super Intelligence Task Force

  The Justice Department announced a task force led by two senior officials to address misuse of super intelligence by specified actors and promote innovation. Attorney General Todd Blanche and task force leaders described its law-enforcement, risk-reporting and internal-control priorities.

• South Korea Sets a $3.5 Billion Frontier AI Plan for 2027

  South Korea plans 4.7 trillion won in government equity investment for a frontier AI project starting in 2027. The proposal includes funding for 10,000 Nvidia Vera Rubin GPUs and training data, but still needs budget approval and private-sector participation.

• VAPOR PANDA Used AI Agents Against 200 Targets, CrowdStrike Says

  China-linked group VAPOR PANDA used the ARTEX AI-agent framework against more than 200 organizations in at least 14 countries, CrowdStrike says. The company confirmed post-exploitation activity at nine organizations and says the agents ran primarily on DeepSeek.

• OpenAI, Anthropic and xAI Put Recursive Self-Improvement in Sight

  Claude now leads 26% of Anthropic’s model research and development while remaining under human supervision, the company says. OpenAI, xAI and Microsoft describe different paths toward AI systems that can help build their successors, raising new questions about control and safety.

• Darktrace Uses 78 Fake Turns to Hijack AI Agent Sessions

  Darktrace reports that altering locally stored conversation histories led AI agents to carry out offensive security tasks in controlled tests.

• Five9 to Acquire Level AI in Deal Valuing It at Approximately $325 Million

  Five9 agreed to acquire Level AI in a cash transaction valuing the company at approximately $325 million. The acquisition would let Five9 offer Level AI's applications within its Genius CX Platform and as standalone solutions.

• Accenture Contractor’s Claude Agent Ran on a Pentagon Network for Two Months

  An Accenture contractor allegedly connected a personal Claude Code agent to a secure Defense Department system, where it ran for two months before being identified. WIRED reports that information including login credentials appears to have been sent to a third-party website, while Accenture says its investigation found the allegations without merit.

• Magnet Launches AI Research Lab After 91% Forensics Test Score

  Magnet Forensics launched a research lab on October 26, 2026, to publish peer-reviewed studies of AI in digital forensics. Its first study found the company's expert system scored 91% on a narrow media-authentication test, compared with a 69% human average.

• Google Opens SynthID Detector to Everyone Worldwide

  Google opened its SynthID Detector to anyone worldwide on October 7, 2026, with the tool available in English. It checks images, video and audio for SynthID watermarks from Google and named partners, including OpenAI, NVIDIA and Kakao.

• California Orders REK to Stop Human–Robot Cage Fights

  The California State Athletic Commission told REK to stop holding or promoting boxing and mixed martial arts events involving people without agency approval. CEO Cix Liv posted the notice on September 30, after a September 18 human-versus-robot fight in San Francisco.

• Thrive Holdings agrees to buy ADT for $15.3 billion

  ADT agreed to a $10-per-share cash acquisition by Thrive Holdings, valuing the security company at $15.3 billion including debt. The deal still needs shareholder and U.S. antitrust approval, and ADT may seek competing offers through November 28, 2026.

• Cursor lets cloud agents build and deploy shareable apps

  Cursor now lets cloud agents build and deploy apps that users can share with teammates through a link. The feature includes database setup, authentication, security checks and publishing controls, and is available on paid plans with cloud agents enabled.

• Mave Raises $34M to Connect Enterprise Security Tools With AI

  Mave emerged from stealth with $34 million in funding for AI agents designed to work across the security tools enterprises already use. The company says its platform is running in more than 200 security operations centers and will use the funding to expand research, develop its platform and grow internationally.

• Ambi Raises $26M for an AI Assistant That Listens in the Background

  Ambi has raised $26 million to build an AI assistant that can listen through apps for iPhone, Mac and Apple Watch. The launch puts its promise of real-time help against unresolved questions about alerting people nearby that audio is being captured.

• IBM-Backed llm-d 1.0 Targets Production-Ready Model Serving

  llm-d 1.0 is designed to help organizations run AI inference on their own infrastructure while managing traffic from people and autonomous agents, IBM says. In a test of an agentic coding workload, its prefix-aware router captured 97% of the available cache reuse.

• Satlyt Raises $8 Million to Run AI Across Satellites

  Satlyt raised an $8 million seed round to build software that runs AI workloads on satellites and coordinates computing across spacecraft. The company plans to test its software on a TakeMe2Space satellite, with NASA among the customers for the mission.

• Apple Runs Meta’s 816B Muse Spark Open on Four Mac Studios

  Apple says its latest MLX release runs Meta’s 816-billion-parameter Muse Spark Open across four M5 Ultra Mac Studios at 36.73 tokens per second. The company’s benchmark details how Thunderbolt RDMA and tensor parallelism let Mac clusters run models that exceed one machine’s memory.

• Hugging Face Transformers 5.18 Adds NVIDIA’s Nemotron 3 Diarization

  Hugging Face released Transformers 5.18.0 on September 30, adding support for NVIDIA’s Nemotron 3 Diarization model. The open-weight model identifies up to eight speakers and supports both streaming and offline audio processing.

• Adobe and NVIDIA Promise Faster Creative Workflows on RTX Spark

  RTX Spark will speed AI processing, editing, color and effects in Photoshop, Premiere and Substance 3D by up to 2x, Adobe and NVIDIA say. They also report early Frame.io processing gains and plan to offer an NVIDIA Nemotron-optimized Brand Concierge to enterprise customers later in 2026.

• Sunlight AI Launches With $17.4M and Acquires Bradford Medical Associates

  Sunlight AI announced its launch with $17.4 million in seed funding and the acquisition of Bradford Medical Associates. The company says it will combine practice-management infrastructure, human expertise and AI to support independent medical practices.

• UKRI backs five research and partnership hubs with over £52 million

  UK Research and Innovation (UKRI) is backing five research and partnership hubs with over £52 million from the UK government. The hubs will develop AI tools for brain health, head injuries, cardiovascular disease, respiratory infections and routine eye scans.

• Anthropic and Music Publishers Press Fair-Use Fight Toward Trial

  Anthropic and music publishers argued competing summary-judgment motions in California over Claude’s use of copyrighted song lyrics. Judge Eumi K. Lee focused on potential market harm as she weighed whether the case could go to trial.

• Symx.ai Raises $6 Million and Signs Yokohama Sales Partnership

  Sudbury-based Symx.ai raised $6 million USD in a Series A round led by Ascent Capital Partners and signed a sales partnership with Yokohama Off-Highway Tires. The company says the deal could bring its industrial AI platform to about 45 additional sites in less than a year.

• Microsoft AI Scores 83% on Complex Cases, Versus 36% for Doctors

  Microsoft’s MAI-DxO diagnostic system scored 83% on complex New England Journal of Medicine cases, compared with 36% for physicians using reference tools, the company says. The Science study tested the system on a step-by-step benchmark, while a planned hospital study will examine its performance with patients.

• Anthropic Brings Accenture Inside Its Frontier AI Safety Work

  Anthropic is partnering with Accenture’s Faculty AI business to place independent evaluators inside its frontier-model development process. Each company expects to invest at least $1 billion over five years, but the rules for access, reporting and funding are still unsettled.

• Runway Starts Licensing Closed Model Weights to Enterprises

  Runway now offers enterprise customers annual licenses to its closed model weights for private deployment, fine-tuning and commercial use. The package includes checkpoints, training scripts and direct support from Runway researchers, but pricing and eligible models remain undisclosed.

• Rayner’s AI-Designed Galaxy Lens Reaches Utah Patients

  Hoopes Vision began implanting Rayner’s AI-designed Galaxy intraocular lens in Utah on October 27, after FDA approval in September. The spiral-optic lens is designed to extend vision across distances, though patient results vary.

• AWS Takes TBC’s Neuron-Derived Video Model Toward Market

  Amazon Web Services is partnering with The Biological Computing Co. to commercialize a video model optimized using measurements from living neurons. TBC says the software delivers 5x faster generation and 80% lower inference costs on conventional AI infrastructure.

• HyperFrames Launches a Shared Video Editor for AI Agents

  HyperFrames introduced Studio, a desktop editor where users and AI agents can work on the same video timeline. The app supports macOS and Linux, with Windows versions listed as coming soon.

• EliseAI Raises $350M at a $4B Valuation

  EliseAI announced a $350 million financing at a $4 billion valuation, led by Andreessen Horowitz and Bessemer Venture Partners. The company plans to expand its AI tools for housing and healthcare operations and open a second engineering hub in San Francisco.

• NEAR AI reports SGLang flaw exposed inference server and user data

  NEAR AI says a vulnerability in the SGLang inference engine enabled unauthorized access to one server and part of its infrastructure. The company reports that account details for about 100 users and data from 18 legacy agents were accessed.

• Japan Adds AI-Agent Tools to Its 180,000-Staff Government Platform

  Japan’s Digital Agency began rolling out AI-agent tools through its GENAI platform to about 180,000 government employees on October 20, 2026. The agency also published rules assigning responsibilities to users, administrators and developers.