Uudised
See voog koondab avalikult kättesaadavaid artikleid kolmandate osapoolte allikatest. Kokkuvõtted genereeritakse automaatselt asjakohaste teemade esiletõstmiseks; need ei ole Ando Kivilaidi originaalreportaaž. Täieliku artikli lugemiseks ava allikas. Valik prioriseerib AI süsteeme, kindlustustehnoloogiat ja ehitajatele olulist inseneriuudist.
- NTT DATA Launches AI Insurance Platform for Carriers(securitybrief.asia)
NTT DATA launched NTT DATA AI for Insurance, an agentic platform built on its AIVista foundation offering prebuilt AI agents for underwriting, claims and distribution workflows in regulated insurance environments.
Google chief scientist Jeff Dean and longtime collaborator Sanjay Ghemawat are departing after 27 years to co-found Discovery Loop, a startup aiming to automate scientific research loops. They are joined by DeepMind's Oriol Vinyals and Google Brain co-founder Quoc Le, with a seed round co-led by Radical Ventures and Khosla Ventures.
- LinkedIn Says It Won't Expand AI Data Centers This Year After Doubling GPU Efficiency(tech.yahoo.com)
LinkedIn announced it will keep GPU investment and data center capacity flat for the fiscal year, an unusual move within Microsoft's portfolio given the company's heavy AI capex elsewhere. LinkedIn said it roughly doubled output from its existing GPUs over six months through better utilization, model distillation and workload reallocation, and will redirect the freed-up capacity toward new features instead of new hardware.
- MS&AD Insurers Deploy AI Image Detection to Combat Claims Fraud in Japan(asiainsurancereview.com)
Mitsui Sumitomo Insurance and Aioi Nissay Dowa Insurance, both part of MS&AD, launched a jointly developed tool with TechUnion that scans claim photos, repair estimates and receipts for signs of AI generation, manipulation or reuse. The insurers say it is the first such capability deployed by domestic Japanese insurers in claims operations, with humans retaining final payout decisions.
- Lockton Expands AI-Enabled SAGE Intelligence Platform Across US Business(reinsurancene.ws)
Insurance broker Lockton widened availability of its Lockton SAGE platform, which combines proprietary data, analytics and AI to consolidate fragmented policy, claims and exposure information, across its Risk Solutions, People Solutions and Reinsurance divisions in the US. The platform originated within Lockton Re before this broader rollout.
Broker Marsh appointed Jessica Fields, a 20-year insurance veteran previously of Aon, Zurich North America and AXA XL, to lead its newly created US Semiconductor and AI Compute Practice in New York. The practice will provide risk advisory and insurance solutions covering physical, operational, digital and financial exposures tied to the rapid buildout of AI computing infrastructure.
Anthropic appointed Mariano-Florentino (Tino) Cuéllar, a former California Supreme Court Justice and ex-president of the Carnegie Endowment, as its first Chief Global Affairs Officer to lead policy and government relations, reporting to president Daniela Amodei.
- Anaconda Acquires AI Security Startup Enkrypt AI(anaconda.com)
Anaconda acquired Enkrypt AI, which has found roughly 143,000 vulnerabilities across 25,000 MCP servers, adding AI security, governance and compliance capabilities to Anaconda's platform.
Intelligence, the company behind Design Arena, raised a $7.9M seed round led by Index Ventures with participation from Conviction, A*, and Valkyrie. The platform crowdsources head-to-head human evaluations of AI-generated media and sells the resulting preference data to frontier AI labs to improve their generative models, reporting 5.3M users and $60M ARR.
- AWS Strikes Multiyear Deal to Embed Vibe-Coding Startup Superblocks in Enterprise Private Clouds(techcrunch.com)
Superblocks announced a multiyear joint marketing agreement with AWS that lets its AI coding tool run inside enterprise customers' private clouds, integrating with Amazon Aurora and Amazon Bedrock while keeping data off external servers. AWS will also help sell Superblocks to enterprises through its Marketplace, reflecting cloud providers positioning as multi-model AI orchestrators.
London-based AI inference chip startup Olix raised a EUR270.5M ($312M) Series B, tripling its valuation to $3.3B roughly six months after its prior round, in what is being called one of Europe's largest semiconductor VC rounds. New investors Fundomo, Arm, and Hudson River Trading joined existing backers including Hummingbird Ventures, Crane and Plural, alongside the UK government's Sovereign AI venture fund.
Alibaba released its largest-ever AI model, the 2.4-trillion-parameter Qwen3.8-Max, a sparse mixture-of-experts multimodal model handling text, images, video and documents. The company claims benchmark scores rivaling or beating Anthropic's Claude and Moonshot's Kimi K3, though open weights and a full model card are still pending.
Guidewire introduced the Agentic Framework in its new Qusar release, letting P&C carriers build, deploy and govern AI agents on the Guidewire Cloud Platform with real-time access to policy, claims and billing data. The release ships pre-built claims and underwriting agents plus Developer Assistants Guidewire says speed configuration work by over 40%.
- Plymouth Rock Launches Home Insurance Quotes Directly Inside ChatGPT(insurancejournal.com)
Plymouth Rock Home Assurance became the first U.S. insurer to let customers get home insurance quotes through natural conversation inside ChatGPT, going live in Massachusetts, New Jersey, Pennsylvania, Connecticut, New York and New Hampshire. The move follows similar AI-distribution experiments by Liberty Mutual, Allstate, Aviva and AXA, amid industry debate over disruption to independent agency commissions.
Cybersecurity startup Horizon3 raised a $250 million Series E at a $2 billion valuation, more than tripling its valuation from 14 months earlier, led by NightDragon and NEA with new backing from Qualcomm and SAIC. The company's NodeZero platform uses autonomous AI agents to continuously penetration-test production systems, positioning itself for what it calls the emerging AI-vs-AI cybersecurity era.
- Commission starts enforcing AI Act rules and new transparency requirements on 2 August(ec.europa.eu)
From 2 August 2026, the European Commission's AI Office and national authorities gained full enforcement and sanctioning powers under the EU AI Act, with new Article 50 transparency rules taking effect requiring chatbots to disclose they are AI and deepfakes or AI-generated content to be labeled. High-risk AI system obligations under Annex III were pushed back to December 2027 following May's Digital Omnibus agreement.
- OpenAI's Unreleased Astra Model Solves Ten Long-Open Math and Theoretical CS Problems(siliconangle.com)
OpenAI said an internal preview of its next model family, Astra, produced machine-checkable Lean proofs for ten problems in mathematics and theoretical computer science that had been open for at least a decade, including the first explicit construction of a non-sofic group. The company published all proof certificates on GitHub, noting the compute cost was roughly $2,000.
- California's AI Transparency Act (SB 942) Takes Effect, Mandating AI Content Watermarking(techtimes.com)
California's SB 942 became operative on August 2, 2026, requiring generative-AI providers with over 1 million California users to offer free AI-detection tools and embed visible and latent provenance disclosures in AI-generated content. Violators face penalties of $5,000 per violation; Midjourney was flagged as a high-profile non-compliant generator.
OpenAI said an internal version of its next model, Astra, produced machine-checkable Lean 4 proofs for ten previously unsolved problems in mathematics and theoretical computer science, including the first explicit construction of a non-sofic group and a new sphere-packing density bound. OpenAI published the proofs and formal verification certificates on GitHub, and said the results cost roughly $2,000 in compute.
- Zywave Expands Apex AI Platform With Summer 2026 Release(fintech.global)
Insurance distribution technology provider Zywave unveiled its Summer 2026 Release for the Zywave Apex AI platform, adding tools for prospecting, renewals, quoting and client management aimed at brokers and agencies. New features include a Producer Agent that integrates with Microsoft Outlook for prospect outreach and scheduling; the platform serves 15,000 global organizations.