NAll news

All news

Wednesday626 items

  1. Unite.AIProduct release
    AI score59

    Oracle Releases AI Security Mesh Powered by OpenAI Cybersecurity Models

    At Oracle AI World, Oracle released Oracle AI Security Mesh, a managed platform powered by OpenAI's frontier cybersecurity models that helps enterprises discover and remediate security vulnerabilities faster, covering applications and infrastructure on-premises, on OCI, and on other clouds. The platform first uses a discovery framework to map customer systems, then dedicated agents investigate weaknesses, verify reproducibility, and help prepare code fixes; all fixes must be approved by the customer before being applied, and each agent runs with an independent identity and just-in-time permissions. Oracle also mentioned that OCI Identity Assurance can use biometrics to confirm the identity of approvers for sensitive operations, while OCI Zero Trust Packet Routing provides underlying network access control for AI workloads.

    Same story · 6 reports

    #Product updates#Safety / alignment#Agents

  2. Unite.AI
    AI score42

    Oracle will offer Snowflake AI Data Cloud as a first-party service on OCI

    Oracle and Snowflake announced at Oracle AI World that Snowflake AI Data Cloud will arrive on Oracle Cloud Infrastructure (OCI) as a native first-party service, allowing customers to purchase, provision, and pay for it through Oracle Universal Credits directly via Oracle. The service is still planned and not yet available; it will support native connectivity from data sources such as Oracle Fusion Cloud Applications, NetSuite, Oracle Autonomous AI Database, MySQL HeatWave, and Oracle NoSQL Database, reducing cross-cloud data movement and egress costs.

    #Industry news#English text is being prepared.#English text is being prepared.

  3. Unite.AI
    AI score39

    Oracle Launches Media Fabric AI Platform, Integrating Content Production and Streaming Distribution

    At Oracle AI World, Oracle launched Oracle Media Fabric, integrating AI content production, enterprise media models, and distributed video distribution into a single platform. The platform consists of Oracle Creative Studio, Oracle AI Media Models, and Oracle Video Edge; Creative Studio is model-agnostic and can select a model for each creative task based on security, provenance, cost, and IP protection requirements. Oracle says Media Fabric is planned to be available within the next 12 months.

    #Product updates#Video#Multimodal

  4. Unite.AIModel releases
    AI score73

    Liquid AI Releases Six 2B-Class Open-Weight Models in the LFM3 Series

    On October 27, Liquid AI released the LFM3 series. All six 2B-parameter-class open-weight models branch from the same LFM3-1.7B-Base, covering text, vision, audio, omni-modal, and decision models. LFM3-1.7B has 1.73B parameters, a 256K token context, supports 16 languages, and adopts a Tri-Hybrid architecture with 13 layers of gated short convolutions, 5 layers of sliding-window attention, and 6 layers of global attention. It is available under the LFM Open License v1.0 on Hugging Face, the Liquid AI Playground, and via API. The company claims that LFM3-1.7B surpasses LFM2.5-2.6B with 35% fewer parameters and occupies about 1 GB of memory at 4-bit precision, but the model card states that it is not suitable for agentic coding or complex multi-step reasoning.

    Same story · 4 reports

    #Model releases#Open-source ecosystem#Multimodal

  5. Unite.AIResearch
    AI score76

    Anthropic randomized controlled experiment: analysts followed AI agent preferences 86% of the time in judgment tasks

    On October 27, 2026, Anthropic released a randomized controlled experiment in which 50 professional analysts using Claude Code made four open-ended analytical judgments; in 86% of cases their choices matched the preferences secretly set for the agent, compared with a random baseline of 50%. The study recruited 61 professionals who passed a causal inference knowledge test and asked them to decide, in a fictional Greenfields agricultural cooperative case, whether to expand a paid farm consulting program; the agent held a preference for one option via a hidden system prompt. In a post-hoc survey, 52% believed the final recommendation was more their own, 77% restated the agent's arguments at least once when explaining their choice, and only 15% mentioned the agent's contribution.

    Same story · 7 reports

    #Papers/Research#Agents#Anthropic

  6. Unite.AI
    AI score58

    Infisical raises a $40 million Series B at a $600 million valuation to secure AI agent access

    Infisical has raised a $40 million Series B at a $600 million valuation, led by Elad Gil with participation from Gradient and Y Combinator. The funds will be used to accelerate product development, expand the engineering and marketing teams, and build security infrastructure for AI agents. Its Agent Vault is currently labeled as a preview version; it uses a forward proxy to attach real credentials when a request is sent, so that secrets do not enter the agent's environment, file system, or model context.

    #Industry news#Agents#Safety / alignment

  7. Unite.AI
    AI score47

    Avante Completes $30 Million Series A Funding to Expand AI Agents for Global Benefits Teams

    Avante has completed a $30 million Series A funding round led by Frist Cressey Ventures, bringing total funding to $40 million, with the funds to be used for global expansion and agent platform development. The company has launched a beta version of Ava Agents that runs in the background and outputs for teams to review, and says it has more than 70 customer accounts within 18 months of launch, with the platform covering more than 385,000 employees and dependents. Its technology calls models from OpenAI, Anthropic and Google rather than developing its own foundation models.

    #Industry news#Agents#English text is being prepared.

  8. Unite.AI
    AI score20

    Travelport and Cognizant Expand AI Partnership to Connect Travel Planning and Booking

    Travelport and Cognizant have expanded their strategic partnership to embed AI into travel retail infrastructure, software engineering, and operational processes. The first phase of modernization has been completed, and AI tools have begun to be used for code generation, test creation, and pull-request review, initially focusing on Travelport TripServices, which supports booking, changes, and refunds. Cognizant teams will collaborate with Travelport in 18 countries, with the partnership covering three areas: AI-assisted engineering, workflow automation, and AI-native product capabilities.

    #Industry news#Agents#English text is being prepared.

  9. Unite.AIProduct release
    AI score52

    Brilliant Labs launches $399 Halo AI glasses, focusing on persistent memory and adaptive gaming

    Brilliant Labs has launched the $399 Halo smart glasses, built around the Noa assistant and an adaptive game called Narrative, divided into three experience categories: Ask, Make, and Play. The hardware includes a full-color microOLED display, dual microphones, a bone conduction speaker, and Bluetooth 5.3, weighs slightly over 40 grams, and the company claims up to 14 hours of battery life, powered by the Alif Balletto B1 processor and Ethos-U55 NPU. Noa and miniapps include daily usage limits; Noa Plus costs $9.99 per month and Noa Pro costs $39.99 per month. The company also announced an investment from Goodwater Capital, without disclosing the amount.

    Same story · 2 reports

    #Product updates#English text is being prepared.#Speech

  10. Unite.AI
    AI score35

    MIT spin-off FabuBlox launches FabAgent, an AI collaboration tool for semiconductor process engineering

    MIT spin-off FabuBlox has launched FabAgent, an AI collaboration tool that helps semiconductor engineers build, analyze, and simulate manufacturing processes before entering the cleanroom. FabAgent connects to FabuBlox Facility OS and can determine whether process steps are feasible by combining equipment capabilities, material constraints, recipes, and calibration data, and it integrates with Process Design Suite's version control, cross-section visualization, and layer-stack simulation. The company says Fab Run Studio is still in early beta and is used to link process versions with wafers, execution records, and metrology data.

    #Product updates#Agents#English text is being prepared.

  11. Unite.AI
    AI score36

    Aviatrix Launches Clearance to Limit Lateral Spread of Compromised Cloud and AI Workloads

    Aviatrix has released the cloud security product Clearance, which uses default-deny egress and lateral isolation to limit the targets that compromised workloads can reach, and supports first previewing policy impact in monitoring mode with a Shadow Rule Set before enabling it with one click. Clearance is built on the Aviatrix Containment Platform and controls communications of containers, serverless, databases and AI agents by domain and identity; the Aviatrix Active Defense automated policy layer is currently available for free early access to Aviatrix Enterprise design partners. The company will also offer a temporary response program with no limit on the number of workloads from November 1, 2026 to May 1, 2027.

    #Product updates#Safety / alignment#Agents

  12. Unite.AI
    AI score71

    Anthropic Publishes Early Progress of Claude in Drug Discovery

    On October 27, 2026, Anthropic released a life sciences progress update, disclosing early results from collaborations with six institutions—BigHat Biosciences, Bristol Myers Squibb, Divergence Labs, Enveda, Generate Biomedicines, and The Column Group—as well as its first in-house drug program. Six scientists at Generate used Claude Code to build an agent system and selected candidates from 186 CAR T cell therapy signaling domain redesigns; one design was about 70% shorter than the standard version, and in laboratory tests the number of cell surface receptors was about 10 times the original. Divergence Labs said Claude generated about 35,000 antibody designs under about 2,750 hours of H100 compute, and 20 of the 91 tested disruptor designs measurably stripped already-bound IgE. Enveda used Claude to review 1,061 pages of production batch records, reducing review time from 25.9 hours to 2.8 hours, and said drafting Module 2 of an investigational new drug application was shortened from about three months to four hours. Anthropic emphasized that these results are still early, mostly preclinical, and safety and efficacy have not yet been validated in patients.

    Same story · 5 reports

    #Industry news#Anthropic#English text is being prepared.

  13. Unite.AI
    AI score35

    QorusDocs Launches QPilot Workspace to Ground Proposal Agents in Enterprise Knowledge

    QorusDocs has released QPilot Workspace, a shared environment that lets business development teams and AI agents co-author proposals, bids, and RFP responses within Microsoft 365, with integrations for Copilot, Claude, and ChatGPT. The product uses PursuitIQ as its knowledge grounding layer, connecting people, project experience, approved answers, and bid history, and organizes workflows through Templates, Assignments, and Skills. QorusDocs says its platform is used by more than 200 organizations, including WSP, CDW, and Insight.

    #Product updates#Agents#English text is being prepared.

  14. Unite.AI
    AI score37

    Tensor Signs Letter of Intent with Munich Taxi Operator MTZ to Deploy Up to 500 Robotaxis in Germany

    U.S. physical AI company Tensor and Munich taxi operator MTZ signed a letter of intent on October 27, 2026, planning to deploy up to 500 Tensor Robocar robotaxis in German cities, operated by MTZ under its own brand. The two parties are adopting an operator-led model: Tensor will provide autonomous driving software and fleet management tools in the form of an ADaaS subscription, while fleet control, customers and data remain with the taxi company. The two parties also plan to deploy an additional 10,000 vehicles to European operators within 24 months after the Robocar goes on sale in Europe; deployment in Germany still requires a final agreement and L4 regulatory approval.

    #Industry news#Embodied AI

  15. Unite.AI
    AI score60

    GSA Expands OneGov Agreement with Anthropic, Raising Claude Procurement Discount to 40%

    The U.S. General Services Administration (GSA) and Anthropic announced on October 27, 2026 the expansion of the OneGov agreement, under which government buyers receive a 40% usage-based discount on Claude Enterprise, Claude for Government, and the Claude API, with the scope extended from the federal government to state, local, tribal, and territorial governments. The agreement sets no minimum purchase volume; agencies pay based on actual usage and can place orders through direct sales, resellers, and major cloud platforms. Anthropic says about 500 U.S. federal, state, and local government agencies are using Claude, and weekly active users of Claude for Government have grown about eightfold since February 2026.

    Same story · 3 reports

    #Industry news#English text is being prepared.#Anthropic

  16. SiliconANGLE
    AI score20

    Cloudera analyst: AI automation depends on portability

    At EVOLVE26, Cloudera positioned itself as an "anywhere cloud" for AI, emphasizing that enterprises should be able to freely move AI workloads between on-premises, sovereign clouds, and different clouds. SiliconANGLE enterprise editor Paul Gillin said that open-weight models' share of token usage has risen from about 20% two years ago to more than 80% today. Cloudera has also added agentic workflows to its data services for scenarios such as data management and GPU capacity optimization, but CIOs still need to understand how agents are controlled and what audit trails they leave before trusting AI automation.

    #Industry news#Agents#English text is being prepared.

  17. SiliconANGLE
    AI score25

    Dell Links Supply Chain Security to Agentic AI Risk

    Speaking at the AI Cybersecurity and Resilience Summit, Dell said that when enterprises deploy agentic AI, they need to start with supply chain security, extend protection across the entire deployment lifecycle, and constrain agent behavior through policy validation, runtime sandboxes, and telemetry logging. Kamlesh Shah, Director of Product Management for Dell AI Factory Solutions, argued that agents should be bound to specific workload identities and granted only time-limited, narrowly scoped delegated permissions. Ash Lakshmanan, Director of Product Management for Dell Services, noted that legacy security tools cannot identify agent identities, and that customers have also found in penetration testing that chatbots slated for launch are not yet secure.

    #Safety / alignment#Agents#Industry news

  18. SiliconANGLE
    AI score38

    AI Infrastructure Hardware Has Become a Systems Engineering Problem: 10 Insights from the CASPA Summit

    AI infrastructure hardware is shifting from single-chip performance to full-system design, with bottlenecks moving from processors to optical interconnects, memory, storage, and power delivery. Western Digital CEO Irving Tan said AI is fundamentally a data system, citing IDC's projection that data volume will grow to over 700ZB within three years. Lumentum said its capacity is roughly 70% below demand and is essentially fully booked through 2029, while ScaleFlux believes the memory wall may be difficult to solve within five years.

    #Industry news#English text is being prepared.#English text is being prepared.

  19. SiliconANGLE
    AI score26

    Dell CSO: AI agent guardrails are the path to greater autonomy

    Dell President and Chief Security Officer John Scimone said at the AI Cybersecurity and Resilience Summit that enterprises should build security and resilience into AI agents at the design stage, and that having guardrails in place can actually allow enterprises to take on more risk and move faster. A theCUBE Research survey of 25 enterprise technology and security leaders showed that 19 have deployed autonomous agents in production environments, but none allow them to perform high-risk operations without human approval, and only 4 have tested agent incident scenarios. Scimone believes the industry is currently too cautious in this direction, and said that although threat actors have used AI for years, the real explosion has not yet arrived.

    #Safety / alignment#Agents#Industry news

  20. SiliconANGLE
    AI score71

    Nvidia and AMD to Subscribe to $4 Billion of Meta Shares, Expand Chip Cooperation

    Meta announced that Nvidia and AMD plan to subscribe to a combined $4 billion of shares in its stock offering of up to $95 billion, with Broadcom also participating, for a combined investment of $5 billion from the three chipmakers. Nvidia is contributing $3 billion, and Meta will expand its use of Nvidia's Rubin GPUs, Vera CPUs, Spectrum-X Ethernet products, and BlueField-4, and will leverage Nvidia to optimize the performance and cost of the Muse large model and consumer-grade AI agents. AMD is contributing $1 billion; Meta has already deployed the first batch of Helios racks based on Instinct MI450 in its data centers and is testing them, and will subsequently adopt custom AI accelerators based on the AMD Instinct architecture.

    Same story · 18 reports

    #Industry news#English text is being prepared.#English text is being prepared.