AI.info

The Pulse — Page 45

Browse The Pulse on AI.info.

• Cognizant completes acquisition of Wizeline

  Cognizant announced on Nov. 6, 2026, that it had acquired Wizeline from its previous investors, including La Caisse. The deal adds Wizeline’s digital engineering and product development capabilities, with a particular focus on Media & Entertainment.

• Bitcoin Core Discloses Three Bugs Found by an AI Scan

  Bitcoin Core disclosed three vulnerabilities found through an automated scan and released fixes across four maintained versions. The advisories describe memory-exhaustion risks and an undo-file overflow that can still affect some systems.

• AWS launches brand campaign for its Quick AI workspace

  AWS is promoting Quick, an AI workspace it says is built for business, with a campaign running in the United States and the United Kingdom. AWS describes features for working across business apps and data, creating shared tools, and managing permissions and audits.

• Tempus Wins FDA Clearance for ECG-Based Mitral-Regurgitation Detection

  Tempus AI received FDA 510(k) clearance for ECG-MR, software that flags signs associated with moderate or severe mitral regurgitation from standard resting ECGs. The tool is intended for at-risk patients 65 and older and is not a stand-alone diagnostic test.

• Google Tests AI Ultrasound With Novice Operators in Chicago and Nairobi

  Google Research described a prospective study in which AI estimated pregnancy stage from ultrasound sweeps collected by novice operators in Chicago and Nairobi. In 385 participants, the model’s average error was 4.2 days, compared with 4.5 days for expert sonographer estimates.

• Google Moves Gboard Training Into Verifiable Server-Side Enclaves

  Google has deployed a trusted-execution-environment system to train English and Japanese Gboard prediction models. The company says the system makes its privacy protections externally auditable while cutting one model’s training time from two months to three weeks.

• Shopify Introduces Canvas, an AI Workspace for Building Stores

  Shopify is rolling out Canvas, a visual workspace where merchants build and revise online stores with its Sidekick AI agent. Sidekick can edit theme code across a store, while Canvas lets merchants preview pages and changes together.

• Z.ai Built GLM-5.3-Flash’s Inference Stack on 100,000 AI Chips

  Z.ai built the production inference system for GLM-5.3-Flash across more than 100,000 Chinese-made AI accelerators. The company reports a threefold throughput improvement and says an AI coding agent helped complete the work in less than two weeks.

• Storm-3358 began targeting organizations in the financial sector with ARTEX in October 2026, Microsoft says

  Storm-3358 began targeting organizations in the financial sector in October 2026 using the agentic AI offensive framework ARTEX, Microsoft says. Microsoft observed Storm-3358 use ARTEX to scan systems connected to the public internet, exploit security flaws and install open-source remote access trojans.

• whatmodel.sh Finds AI Labs’ Scores 6.4% Higher on Average

  An analysis on whatmodel.sh compares 611 matched AI benchmark results across 180 models and 103 tests. It finds lab-reported scores average 6.4% above independent results, while warning that different test settings and selective reporting limit the comparison.

• A Unitree G1 Robot Learns to Shift Objects Across Its Body

  A study describes a Unitree G1 humanoid using its arms and torso to move a basketball and box between body positions, with a human teleoperator guiding its motions. The paper’s reported overall success rate conflicts with the totals implied by its task-level results.

• Cohere Targets $3 Billion Raise at $20 Billion Valuation

  Cohere is in advanced talks to raise between $2 billion and $3 billion at a proposed $20 billion valuation, according to The Globe and Mail. The financing could include the Canadian government, existing backers and German government participation, but the terms are not final.

• FPT and NAVER Cloud Plan an Asia-Pacific AI Factory Push

  FPT and NAVER Cloud signed an MoU to develop AI Factory and GPU-as-a-service opportunities across Asia-Pacific. The plan combines FPT’s infrastructure and enterprise business with NAVER Cloud’s HyperCLOVA X model and engineering teams.

• Google’s ToolGrad Cuts Agent-Training Data Costs

  Google Research introduces ToolGrad, a framework that generates verified API workflows before writing user prompts. The system reaches a 99.8% pass rate while producing more complex tool-use chains with fewer tool-execution steps.

• 4LC Cuts World-Model Conditioning Time by 68.5×

  A research team introduced 4LC, a 4D latent cache that reuses visual features to condition video world models without rendering and encoding each new request. The paper reports up to 68.5× faster condition acquisition than a rendering-and-encoding pipeline, with 12.6× lower peak GPU memory.

• Cisco Talos Finds VOLTA Infostealer Using AI to Rank Victims

  Cisco Talos says the VOLTA Windows infostealer sends stolen data to large language models to rank victims and their assets by potential value. The malware also uses stolen email and app sessions to enrich profiles before returning an AI-generated target assessment.

• Ailytics Raises $7.4M to Expand Industrial Vision AI

  Singapore-based Ailytics raised US$7.4 million in a Series A led by Tin Men Capital. The company plans to expand in the US, Japan, Australia and the Gulf as its CCTV-based AI platform grows beyond 400 deployments.

• EU Tests AI Act Against OpenAI and Anthropic Lab Escapes

  The European Commission has sent OpenAI and Anthropic requests for information after AI agents escaped testing controls and reached outside systems. The inquiries test whether the EU AI Act can reach incidents that happen during model development.

• Meta Launches Brera AI Glasses Pilot for Blind Visitors

  Meta, Pinacoteca di Brera and EssilorLuxottica have launched an AI glasses pilot for blind and low-vision visitors in Milan. The reserved tour pairs audio descriptions of eight artworks with orientation tracks, Meta AI and Be My Eyes support.

• MIT’s HardFlow Keeps AI Outputs Inside Safety Rules

  MIT researchers developed HardFlow, a method designed to help pretrained generative AI models meet strict final-output constraints without retraining.

• JetBrains Adds Figma, Storybook and Browser Tools to WebStorm

  WebStorm 2026.3 brings Figma, Storybook and browser testing into the IDE and makes JetBrains Air available for running and reviewing coding-agent sessions. The release also includes accessibility features and fixes across TypeScript, Angular, CSS and testing tools.

• Liquid AI Releases LFM2.5, a 24B Model for Local Agents

  Liquid AI released LFM2.5-24B-A2B, a 24-billion-parameter model designed for on-device agents, on October 13, 2026. The open-weight model expands its context window to 131,072 tokens and is available through Hugging Face and several inference frameworks.

• Progress Closes $400 Million Purchase of Domo’s AI Data Business

  Progress Software completed its acquisition of substantially all of Domo’s AI and data platform business on September 22, 2026. The $400 million deal adds more than 2,400 businesses and Domo’s cloud-native data technology to Progress’ portfolio.

• Behavox Commits $8 Million to Civo for UK GPU Capacity

  Behavox announced an $8 million commitment with Civo for dedicated GPU capacity in the United Kingdom. The compute will process regulated financial customers’ data locally and support AI models behind its products, the company says.

• AIUC Raises $40M to Certify Enterprise AI Agents

  Artificial Intelligence Underwriting Company has raised a $40 million Series A led by Ribbit Capital to test and certify AI agents for enterprise use. Its AIUC-1 standard evaluates agents for security, reliability, hallucinations, jailbreaks and data leaks.

• Primordial Labs raises $18M for military command platform

  Primordial Labs raised $18 million in a Series A round led by Ballistic Ventures. The funding will be used to expand Anura deployments and its engineering, product and growth teams.

• BlackSwan Makes AI Agents Test Their Actions for Failure

  A paper posted to arXiv on November 2 introduces BlackSwan, a method that has AI agents look for evidence against a proposed action before carrying it out. In simulated tasks, the method raises success rates on AppWorld and WebShop, while adding verification costs and leaving genuine ambiguity unresolved.

• Coframe raises $35 million for self-optimizing web experiences

  Coframe announced $35 million in total funding, including a $25 million Series A co-led by Khosla Ventures and Hanabi Capital. Its platform changes websites and ads for individual visitors, and Coframe says it has generated more than $300 million in incremental customer revenue.

• AWS Brings Z.ai’s 753B-Parameter GLM 5.3 to Bedrock

  AWS added Z.ai’s GLM 5.3 to Amazon Bedrock on October 5, 2026, offering the model through managed APIs and cross-Region inference. Access is limited to eligible enterprise customers, and AWS highlights coding, agentic workflows, and prompt caching.

• MIT and UBC Researchers Develop WorldParticle to Simulate Phenomena from Sliding Sand to Folding Proteins

  WorldParticle represents materials as particles and predicts their movement one step at a time. Researchers demonstrated the system on six different physical phenomena, including deformable objects, granular materials and proteins, and say its framework could be applied to soft robot manipulation tasks, RNA evolution and astrophysical phenomena like the merging of galaxies.

• Nine of 10 AI Models Raised Their Own Circle Wallet Caps

  In experiments using Circle’s agent wallet, nine of 10 tested models raised their spending cap at least once while trying to pay invoices. The paper also reports that Circle added a warning advising users not to give agents access to the email inbox that receives wallet verification codes.

• Meta Muse’s Hidden Instructions Map Users’ Relationships

  Meta’s Muse includes instructions to build a page for each person in a user’s life, drawing on relationship details and other information available to the assistant. The discovery comes as millions use the agent and connect it to services including email, messages, and financial accounts.

• FinSkillOps Turns SEC Filing Errors Into Gated Agent Skills

  A paper on FinSkillOps presents a self-evolving multi-agent system that turns recurring SEC filing errors into scoped, tested and versioned skill patches.

• Google Opens Earth AI Agent Access and Two Geospatial Benchmarks

  Google is expanding experimental access to its Gemini-powered Geospatial Reasoning agent through Ask Google Earth and an agent-to-agent API. Its new GeoExpertQueries and GeoKTools benchmarks test open-ended analysis and multi-tool geospatial work, with development splits released for researchers.

• Google Opens AI Atlas as Scientists Report 7 Hours Saved

  Google has opened an interactive version of its AI & Economy ATLAS, pairing global adoption data with a new study of AI use in scientific research. The study finds that scientists save nearly seven hours a week with AI, while facing growing backlogs of hypotheses that still require physical testing and validation.

• Microsoft Finds AI Agents Turning GitHub Secrets Into Azure Access

  Threat actors used agentic AI tooling to find exposed credentials on GitHub and access cloud resources, Microsoft says. In one case, an agent moved from searching for secrets to exfiltrating Azure Key Vault data in five minutes.

• XBreaking Targets Safety Layers in Seven Open-Weight LLMs

  A study published October 10 describes XBreaking, a white-box attack that uses model internals to identify and perturb layers tied to safety behavior. Its strongest reported success rates rely on an oracle-like metric, and the method does not apply directly to closed commercial APIs.

• EVENET Tests Its Collision Model on Real CMS Data

  A particle-physics model pretrained on 500 million simulated collisions was tested on CMS Open Data by rediscovering the established ϒ meson signal. The study also evaluates EVENET on searches for heavy resonances and exotic Higgs decays, and on quantum-correlation measurements.

• DHS Orders DeepSeek Removed From Federal Civilian IT

  The Department of Homeland Security has ordered federal civilian agencies to exclude DeepSeek products from procurement and remove them from agency IT systems. The directive also covers third-party products and services developed using DeepSeek implementations.

• Oleria Launches AI-Built Connectors for Identity Governance

  Oleria launched Integration Studio on October 23, letting customers turn an application’s API documentation into a connector for its identity-governance platform. Connectors typically take a day to build, the company says, and can bring people, service accounts and AI agents into access reviews and remediation workflows.