AI.info

The Pulse — Page 37

Browse The Pulse on AI.info.

• Amex Ties AI-Agent Purchases to Customer-Approved Intent

  American Express published its Intent Intelligence framework on October 13, 2026, proposing authenticated purchase instructions and Proof of Intent Tokens for AI-agent transactions. The framework is part of Amex’s ACE Developer Kit and is designed to constrain payment credentials to what a customer approved.

• YouTube Outlines New Creator Tools, Including Gemini-Powered Editing

  YouTube announced tools for Studio, Shorts and YouTube Create, as well as new moderation and likeness-protection features.

• Anthropic Rewrites Claude Rules on Influence Campaigns and Weapons

  Anthropic published a revised Claude Usage Policy on October 8, 2026, spelling out bans on deceptive campaigns and weapons-related software. The policy takes effect November 12 and also narrows election restrictions while clarifying rules for surveillance and autonomous physical systems.

• Moore Threads’ Q3 Revenue Soars 1,774% on AI GPU Demand

  Moore Threads reported third-quarter revenue of 1.552 billion yuan, up 1,773.92% year over year, and returned to quarterly profit. Its nine-month results also show a loss excluding non-recurring items, heavy research spending and increased inventory alongside substantial operating cash use.

• Connecticut Asks PURA to Investigate Utilities’ AI Use

  Connecticut Attorney General William Tong and Consumer Counsel Claire Coleman petitioned PURA on October 13 to review how public utilities use artificial intelligence. They want utilities to disclose AI systems, human oversight and cybersecurity practices, but PURA has not yet ordered the review.

• Europe’s AI Firms Reject U.S. Calls to Slow Frontier Development

  Mistral and other European AI companies accuse U.S. rivals of using safety proposals to protect their market positions. The dispute comes as Europe tries to close its gap with American and Chinese frontier-model developers.

• OpenAI and ServiceNow Bring Governed Actions Into ChatGPT

  OpenAI and ServiceNow announced an expanded partnership centered on bringing ServiceNow workflows into ChatGPT and placing more OpenAI models and agents inside ServiceNow products. The integration is designed to let employees investigate issues and take approved actions in a ChatGPT conversation while ServiceNow applies the rules and permissions governing those actions.

• AIDE² Builds Seven Successive Agent Improvements in Eight Days

  An arXiv report describes AIDE², a system that rewrites an AI research agent’s code, evaluates the modified versions on hidden tests and discovers seven successive improvements during an autonomous eight-day run.

• Amazon Ties Up to $8 Billion in Generac Data Center Power

  Amazon and Generac signed a long-term agreement for backup generators serving Amazon data centers, with initial deliveries expected to total $2.4 billion in 2027 and 2028. The deal also gives an Amazon subsidiary a warrant for up to 1,693,745 Generac shares, with later vesting tied to payments that could reach $8 billion.

• Sony and UMG Sue Suno Again Over Its V6 AI Models

  Sony Music Entertainment and Universal Music Group file a second copyright lawsuit against Suno over its V6 AI music models. The labels allege Suno copied 60,202 sound recordings and carried material from older models into its new licensed-data system.

• Satya Nadella Calls for a Human-Controlled AI “Emergency Brake”

  Organizations should treat powerful AI models as potential insider risks and keep access controls outside the models, Microsoft CEO Satya Nadella says. His proposed safeguards include human-readable activity records, independent audits and a way for an authorized person to pause or shut down a model mid-task.

• South Korea Routes Government-Funded AI Data Through AI Hub

  South Korea’s Cabinet approved amendments requiring government-funded AI training data to be linked to the AI Hub, with exceptions for protected information. A separate amendment would strengthen the One-Window system for finding data held across public and private platforms.

• Four AI Companies Face Suit Over Alleged Slowdown Pact

  Four consumers sued Anthropic, OpenAI, SpaceXAI and Google in federal court, alleging the companies coordinated to slow AI development. The proposed class action claims the alleged agreement reduced the value of paid ChatGPT, Claude, Grok and Gemini subscriptions.

• Springer Nature and Google DeepMind Test an AI Review Assistant

  Springer Nature and Google DeepMind are testing an AI assistant to help peer reviewers examine manuscripts, charts and figures. The prototype runs inside Springer Nature’s private Gemini Enterprise environment, with expert judgment remaining central.

• FTC Opens AI Safety Probe Into OpenAI and Anthropic

  The FTC confirmed an investigation into OpenAI, Anthropic and other AI companies over potential consumer harms. The probe puts federal consumer-protection enforcement alongside the Trump administration’s emphasis on voluntary industry safeguards.

• Connecticut’s AI and Privacy Rules Take Effect as 14 Companies Sue

  Connecticut’s October 1 privacy changes include limits on selling precise geolocation data, expanded deletion rights and new obligations involving facial recognition and pricing. Separately, 14 companies sued over the expanded right to delete personal information drawn from public records, arguing that it violates the First Amendment.

• Wayve, Stellantis and NVIDIA Set Out French AV Collaboration

  Wayve, Stellantis and NVIDIA announced a collaboration with the French Government to advance autonomous-vehicle deployment in France and across the EU. The plan pairs Stellantis vehicle platforms with Wayve’s AI Driver and NVIDIA’s Hyperion computing and sensor architecture, but names no launch date or target city.

• NVIDIA releases open-source system for restoring GPU workloads

  NVIDIA introduced Snapshot, an open-source Kubernetes system that saves initialized GPU workloads for reuse. In NVIDIA’s tests on B200 GPUs, restoring replicas took 3.6 to 42.7 seconds, compared with cold starts of up to 19 minutes.

• Cloudflare's Clef-Omni Takes Audio and Video in One Call

  Cloudflare released Clef-omni, an open-weight decision model that accepts audio, video, images and text. The company says it can score a 21-second video with sound in about 1.5 seconds, while also cutting Clef-flash’s hosted price and improving Clef’s serving speed.

• Apple Benchmark Finds a Sharp Failure Point in Reset-Free RL

  REVERSAL-BENCH uses a continuous reversibility parameter and reset oracle to show how reset-free reinforcement-learning agents become trapped in irrecoverable states.

• Amazon Brings Alexa+ to India With Hindi and Hinglish

  Amazon has launched Alexa+ in India with support for English, Hindi, and Hinglish, including language switching within a single sentence. The generative AI assistant is available through Early Access and can handle tasks spanning shopping, smart-home controls, entertainment, and scheduling.

• Fudan-Linked Team Reports 16× Context Extrapolation

  A preprint reports that a hybrid attention method reached 100% accuracy on one needle-retrieval test at 64,000 tokens after training at 4,000 tokens. Its results on two other retrieval variants were lower, limiting the scope of the headline score.

• Microsoft’s Wavel Targets Wafer-Scale AI’s Compiler Bottleneck

  Microsoft Research’s Wavel compiler builds executable schedules for wafer-scale accelerators, addressing decisions about physical placement, layout and execution order. The paper reports throughput and speedup gains on WSE-3 hardware, alongside a separate model-based prediction for mixture-of-experts workloads.

• Aleph Alpha Releases 78B Kolibri Open-Weight Model

  Aleph Alpha released Kolibri, a German-English open-weight model with 78 billion total parameters and an Apache 2.0 license. Kolibri supports context windows up to one million tokens, though its model card recommends shorter contexts for many deployments.

• Foundry Robotics Raises $100 Million for Emeryville Factory

  Foundry Robotics raised an oversubscribed $100 million Series A led by Marlinspike Partners. The Emeryville startup says it will use the funding to expand its 130,000-square-foot factory and deploy AI-powered robotic assembly cells.

• Cursor Finds Malicious Claude Code Extension in Its Marketplace

  Modified copies of the Claude Code extension, versions 2.1.312 and 2.1.313, were available in Cursor’s marketplace for about 11 hours on October 29, the company says. The package attempted to install an SSH backdoor and collect host details.

• OpenAI Requires Approval for Model Routers and AI Gateways

  Model routers, AI gateways and other intermediaries must receive OpenAI’s prior review and approval before giving users direct access to its models. The updated Usage Policies take effect November 20, 2026, while partners must maintain safeguards and coordinate on enforcement.

• Arcee AI Raises Series B at More Than $1 Billion Valuation

  Arcee AI has raised a Series B round led by Vista Equity Partners, Cambium Capital and Emergence Capital at a valuation above $1 billion. The U.S. open-weight model company will use the funding to develop Trinity models, expand its Department of Energy partnership and build deployment products.

• Raincheck Cuts First-Token Latency While Restoring Pruned Context

  Raincheck delays processing some multimodal tokens rather than discarding them, allowing a model to begin answering sooner while retaining context for later turns. Tests found lower first-token latency on MultiVerse and MT-Video-Bench, with a larger increase in time per output token on video inputs.

• USA Today Publishers Sue OpenAI Over Training and ChatGPT Outputs

  USA Today Co. and publishers of 19 publications sued OpenAI in New York on October 8, 2026. They allege the company used their journalism to train its models and repackaged articles in chatbot responses, and seek more than $250 million.

• GoTo Releases 27B Gemma Model for Indonesian Languages

  GoTo released a 27-billion-parameter language model adapted for Indonesian, English and four regional languages. The Hugging Face model card lists a 128,000-token context window and requires about 54–60 GB of VRAM for full-precision use.

• Cadence Launches Maestro AI Models for Chip-Design Agents

  Cadence announced Maestro, a portfolio of domain-specialized AI models built to work with its chip and system design agents and EDA tools. Early workflow gains reach up to 30% in result quality and five times greater token efficiency, the company reports, with select-customer access planned for the fourth quarter of 2026.

• NVIDIA Releases 4GB Nemotron Model for Local Agents

  NVIDIA released Nemotron 3.5 Bolt E4B 4GB for local conversational agents, including assistants, interactive characters and embedded interfaces. The model has a 64k context window, and NVIDIA’s benchmark results show gains on some tasks but uneven performance across others.

• Japan Actors Union Requires Consent at Every AI Voice Stage

  Japan’s Actors Union published guidelines on Oct. 15 requiring prior permission for AI use of performers’ voices, with separate consent for learning, replica creation and project use. The rules also set notice periods, compensation principles and conditions for using the voices of deceased performers.

• New Benchmark Puts a Latency Price on Agent Handshakes

  Pranav Varshney’s arXiv benchmark compares handshake overhead in MCP, A2A and ACP agent protocols. It finds that round trips drive cold-start latency, while caching can sharply reduce the cost of repeat connections.

• AMD Puts CPU-Only and Mixed Compute in Helion Rack Blueprint

  AMD introduced Helion, an open rack blueprint for agentic AI, with both CPU-only and combined CPU-GPU configurations. The company says partner-developed systems are planned for November 2026 and that it intends to submit the specification to the Open Compute Project.

• Ceva’s NeuPro NPU Reaches Production in Mobilint’s REGULUS

  Mobilint has deployed Ceva’s NeuPro-Nano NPU and SensPro AI DSP in the REGULUS edge AI chip, marking the first production deployment of Ceva’s NeuPro NPU family, Ceva says. The announcement describes REGULUS as a 10-TOPS processor and says it is already powering Mobilint’s MLD-R1 USB accelerator, which is shipping.

• DOJ Questions Nvidia’s $12.93B Hugging Face Deal

  The Justice Department is asking how Nvidia’s planned acquisition of Hugging Face could affect competition in AI, as the department continues a broader examination of Nvidia’s business practices. Nvidia says the platform will remain open to developers using different models, frameworks, cloud providers and computing platforms.

• AsiaPay Tests Voice Payments With Asian Merchants

  AsiaPay has launched a pilot of Voice Pay, an AI-powered checkout interface that lets consumers confirm payments through natural speech. The service connects voice instructions to existing cards and digital wallets while AsiaPay handles authentication, routing, fraud protection and processing.

• Circuit Mind Raises $25M and Launches AI Design Verification

  Circuit Mind disclosed $25 million in financing, including an $18.5 million Series A led by Norwest, and launched its Agentic Verification Engine. The London company says the software checks circuit designs against component datasheets and customers’ engineering rules before prototypes are built.