AI.info

The Pulse — Page 16

Browse The Pulse on AI.info.

• An AI Market Correction Could Reach Global Markets, the IMF Says

  The IMF warned on October 13 that a reversal in AI investment or profits could spill into wider financial markets. Its Global Financial Stability Report also flags rising bond yields, cyber risks and growing borrowing by hyperscalers.

• Progress Adds a Smart Agent to Its Enterprise RAG Platform

  Progress added a Smart Agent, a Microsoft Teams app and a WordPress plugin to its Agentic RAG platform. The agent can split questions into steps and retrieve information from indexed content, connected business systems and the web.

• Cirrascale launches production platform for multi-accelerator inference

  Cirrascale Cloud Services has released a production inference platform that routes models across NVIDIA, AMD, Qualcomm, and Tenstorrent accelerators. The platform adds private model hosting, fine-tuning, governance controls, spend management, and support for enterprise AI applications.

• New Mexico Supreme Court Fines Lawyer $5,000 Over ChatGPT Brief

  The New Mexico Supreme Court held Santa Fe lawyer Stephen Aarons in direct contempt after his murder-appeal brief included fabricated witnesses, false testimony and distorted legal authorities generated with ChatGPT. The court struck the filings, barred Aarons from appearing before it and referred him to the state Disciplinary Board.

• Stanford Study Finds 44% of Tested LLM Claims Expired

  A Stanford study finds that 44% of 50 tested claims about large language models had expired by September 2026. Owen Queen and James Zou compared findings across 20 models released by eight developers.

• Chrome’s Gmail-Sourced Autofill Requires Personal Intelligence

  Chrome is rolling out Gemini-powered autofill suggestions drawn from Gmail for users who have Personal Intelligence enabled in Gemini in Chrome. The feature is coming to users in the United States on desktop, Android and iOS, with options to remove suggestions or disable enhanced autofill in Chrome settings.

• Voyager Releases vork 0.1.0 as a Research Preview Where AI Agents Write 3D Scenes as Code

  vork 0.1.0 is a source-available, headless 3D and VFX engine from Voyager that lets AI agents create scenes as code and render them without a graphical editor. The research preview supports recent NVIDIA GPUs and Apple Silicon but ignores fog and participating volumes.

• Attackers Hijack at Least 17 Exposed Cairn AI Harnesses

  Attackers inserted malicious tasks into at least 17 exposed Cairn panels, directing AI agents to target the systems running them. NorForge describes the campaigns as scripted, with evidence of possible human intervention but no conclusive proof that AI agents orchestrated the intrusions.

• Kyutai’s Voice of Reason Scores 70.6% on GSM8K

  Kyutai’s GLM-4-Voice of Reason is a speech-to-speech model that reasons before speaking and scores 0.706 on a 1,310-item written-channel GSM8K evaluation.

• Google says Gemini Robotics 2.5 raises task success 30% to 100% over earlier baselines

  Google DeepMind introduced Gemini Robotics 2.5, a vision-language-action model it says performs more tasks and learns skills faster. The company reported higher task success in evaluations and a 96% success rate in a factory proof of concept with Apptronik and Mercedes-Benz.

• Mistral Releases Timestral Forecasting Models as Small as 5M Parameters

  Mistral released Timestral, a family of four time-series forecasting models ranging from 5 million to 209 million parameters. The company says the models can make probabilistic forecasts on CPU and rank first on three public benchmarks among pretrained, zero-shot systems.

• Volantis Raises $88M for Photonic AI Memory System

  Volantis announced an $88 million Series A to develop a photonic system for AI inference, with its A-1 design targeting models above 20 trillion parameters. The company plans to deliver its first integrated inference engines to customers in 2027.

• NVIDIA Expands RTX Spark Software Ahead of Oct. 16 Launch

  NVIDIA’s RTX Spark laptops will be available October 16, while compact desktops are scheduled to go on sale in November. The company’s October 13 update details new agent features, an open model, a local-compute router and changes to creative software.

• ITOKI Launches Three AI Agents for Workplace Management

  ITOKI began offering three AI agents for workplace and facilities management on Nov. 4, 2026. The tools cover office portfolio analysis, move and renovation projects, and unused meeting-room and desk reservations.

• Anthropic Donates Agent Skills Standard to Linux Foundation’s AAIF

  Anthropic is donating its Agent Skills standard to the Linux Foundation’s Agentic AI Foundation. The company says more than 70 platforms support the format, which packages task instructions and resources in folders built around a SKILL.md file.

• VAST Data Unveils Hardware-Isolated Runtime for Confidential AI

  VAST Data introduced DataEnclave, a confidential AI runtime built on NVIDIA Confidential Computing. The system is designed to let proprietary models process sensitive enterprise data while keeping both model weights and customer information protected from infrastructure operators.

• Claude Traces NIST’s Prime-Curve Seeds to Unix Timestamps

  Claude traced the origins of NIST’s five prime-field ECDSA curve seeds to a 1990s random-number generator initialized with Unix timestamps, Anthropic says. The finding explains those seeds’ provenance but leaves the B-163 seed and several curve base points unresolved.

• AMD Clears $1 Trillion as Investors Bet on Its AI Chip Growth

  Advanced Micro Devices crossed a $1 trillion market valuation on September 21, 2026, as its shares reached a record high on renewed demand for AI computing. The milestone makes AMD the fourth U.S. chipmaker to surpass that threshold, following Nvidia, Broadcom and Micron.

• Reddit Shifts Toward Logged-In Access as Search Referrals Fall

  Reddit plans to test a primarily logged-in experience outside the United States after search referrals fell by more than one million per week on average in October. CEO Steve Huffman says the change is meant to improve personalization, ad targeting and control over access to Reddit’s data, while leaving open how much content logged-out visitors can see.

• Hugging Face Launches ProtoStar to Make Public Science AI-Ready

  Hugging Face and university and research partners launched Project ProtoStar to make publicly funded datasets, models and code easier to discover and reuse. Carnegie Mellon University will host the first challenge, testing AI forecasts for 210 U.S. cities from November 2 to 13, 2026.

• Glow Finds 13,000 AI-Agent Screenshots on Public GitHub

  Glow Security says AI coding agents posted more than 13,000 internal screenshots to public GitHub repositories across more than 300 organizations. The images include billing records, financial consoles and unreleased product features.

• Anthropic Launches Claude Sonnet 5.5, Claiming 30% Faster Output

  Anthropic released Claude Sonnet 5.5 on September 28, 2026, positioning it as a faster, lower-cost model for everyday work and coding. The model generates output more than 30% faster than Sonnet 5 and can cost up to 30% less per task, the company says, while adding new cybersecurity safeguards.

• Crusoe Raises $3.9 Billion at $30.9 Billion Valuation

  Crusoe has closed a $3.9 billion Series F funding round to expand its vertically integrated AI infrastructure business. The Denver company says the financing will support large data-center campuses, modular Crusoe Spark units and the growth of Crusoe Cloud.

• Magentic Raises $18 Million for Industrial AI Workers

  Magentic has raised an $18 million Series A led by Felicis to expand AI workers for procurement and supply-chain operations at large manufacturers. The London- and New York-based company says its agents already manage more than one million orders for one customer and have helped another save $4 million.

• Taiwantrade Launches ASTRA AI Assistant for Buyers Seeking Taiwanese Suppliers

  Taiwantrade says its new ASTRA assistant uses conversational requests, supplier data and AI-powered search to help international buyers find Taiwanese products and suppliers. The platform describes three features: semantic search, multi-round recommendations and automated requests for quotation.

• Arda Raises $160M Series A for Manufacturing AI Software

  Arda’s $160 million Series A brings total funding to $232 million and values the manufacturing AI startup at $1.5 billion. Its software is already used at Anduril and is designed to help engineers update assembly lines when product designs change.

• AI Is Speeding Up Cyberattacks, Australia’s ASD Says

  AI is helping attackers find vulnerabilities faster and scale phishing, Australia’s Signals Directorate says, as its cyber centre responded to more than 1,500 incidents in 2025–26. The agency urges organisations to patch systems quickly, retire legacy technology and strengthen basic cyber controls.

• Google Agents Reach 42.1% Hit Rate on Alpha-Thrombin Aptamers

  AI agents raised confirmed aptamer hit rates against alpha-thrombin from zero to 42.1% across three experimental rounds, Google Research says. The company also introduced VERID, a digital twin for checking and tracking robotic lab protocols, while keeping human approval in the workflow.

• SolarWinds Launches Avior, a Goal-Driven AI Agent for IT Teams

  SolarWinds launched Avior on October 14, 2026, pitching it as a goal-oriented agent for managing IT operations across hybrid environments. Its initial access focuses on SolarWinds Observability Self-Hosted and network-engineering workflows, with higher-risk actions subject to approval controls.

• Headline Closes $400M Fund for European Seed and Series A Startups

  Headline has closed its $400 million EU VIII fund for European Seed and Series A companies, with a focus on AI. The firm’s European portfolio includes Mistral and Black Forest Labs, and it says its Global Growth Fund can support rounds of $100 million or more.

• NVIDIA Releases Five Nemotron 3.5 Retrieval Models

  NVIDIA released five Nemotron 3.5 models for searching text and images, reranking results and extracting structure from documents. The company says the models cut agent search costs and lead selected retrieval benchmarks.

• Guangzhou and Shenzhen Fine GEO Firms Over Fake AI Search Rankings

  Regulators in Guangzhou and Shenzhen fined GEO service providers for distributing fabricated commercial rankings and scores that could influence AI-generated answers. The reported penalties were 120,000 yuan and 50,000 yuan, respectively, with the cases described as false promotion under competition rules.

• Meituan’s LongCat Publishes a Framework for Training Data Agents

  Meituan’s LongCat Interaction Team published Data World, an open framework for creating synthetic data-analysis environments, tasks and grading rubrics for AI agents. The paper reports 2,135 training tasks and gains for a model trained on them across four benchmarks, while noting limits to what the experiments establish.

• Nanbeige Releases 33B Open Model With 512K Context

  Nanbeige has released Nanbeige4.3-33B, an Apache 2.0 model with a 524,288-token context window. Its model card reports strong scores on agentic coding and workplace-task benchmarks, while describing the model as a teacher for a planned smaller model.

• Cognition Crosses $1B in Annualized Revenue Run Rate, Company Says

  Cognition crossed $1 billion in annualized revenue run rate on September 25, 2026, according to the company. The company’s latest announcement names Devin customers but gives no revenue breakdown or calculation method.

• Avante Raises $30M to Put AI Agents on Benefits Teams

  Avante raised a $30 million Series A led by Frist Cressey Ventures, bringing its total funding to $40 million. The company plans to expand internationally and build AI agents for benefits teams, while acknowledging the risks of inaccurate answers about coverage.

• Drone Strike Shuts Yandex’s Sasovo Data Center

  A drone strike and fire halted operations at Yandex’s Sasovo data center in Russia’s Ryazan region on October 8. The company said that some services could be unavailable from its initial Thursday-morning statement, while the extent of the damage and the condition of equipment remain uncertain.

• Check Point Finds Two LiteLLM Flaws That Could Expose AI Gateways

  Check Point Research disclosed two LiteLLM vulnerabilities that could let authenticated users take over gateway settings or redirect AI requests to steal provider credentials. LiteLLM fixed the flaws in version 1.83.9, and Check Point urged organizations to check direct and embedded deployments.

• OpenAI Acquires Orchestra Bio to Build Drug-Development Agents

  OpenAI has acquired Orchestra Bio, whose AI agents help biopharma teams coordinate drug-development programs. The company plans to combine Orchestra’s platform with its Rosalind life-sciences tools.

• Trimble announces AI capabilities for construction platform

  Trimble announced AI capabilities for its Trimble Connect construction platform, spanning design, building and operations. Some capabilities are available now, while others remain in development or a customer preview program.