GPT-5.6 delayed; Anthropic fights distillation

Anthropic / Claude ecosystem

Anthropic is testing desktop-like Claude Cowork for mobile

Anthropic is testing a mobile version of its Claude Cowork agent that would let users start and monitor tasks from a phone while heavy processing runs on a desktop.

Frontier model providers

Previewing GPT-5.6 Sol: a next-generation model

OpenAI previewed GPT-5.6 Sol, a next-generation model family with Terra and Luna variants aimed at improved performance and lower cost, ahead of a controlled rollout.

AI developer tooling & infrastructure

Mindstone Launches Rebel: Local-First AI Agent OS with Shared Memory

Mindstone released Rebel, an open-source local-first AI agent operating system with tiered memory and multi-model orchestration to avoid vendor lock-in.

Santander AI Lab releases open-source Autoguardrails on Cloudflare Workers

Santander AI Lab open-sourced Autoguardrails, a production-ready LLM guardrail that runs at the edge on Cloudflare Workers and detects jailbreak attempts.

Cloud & platform providers

OpenAI GPT, GPT OSS, and NVIDIA Nemotron models on Bedrock receive FedRAMP High and DoD IL-4/5 approval

These models are now available in AWS GovCloud with FedRAMP High and DoD IL-4/5 approvals, enabling government and defense use.

Amazon EC2 C7a instances now available in Asia Pacific (Singapore) Region

AWS's C7a instances with up to 50% higher performance than C6a are now available in the Singapore region, suited for compute-intensive workloads.

Cloudflare Launches Self-Managed OAuth for All Customers

Cloudflare now allows all customers to create and manage OAuth clients for delegated API access, upgrading to Hydra 2.X for better performance.

AI policy, regulation & governance

Anthropic Urges Congress to Crack Down on AI Distillation By Chinese Rivals

Anthropic has formally asked the US Congress to legislate against adversarial distillation, which it says allows Chinese rivals to replicate frontier AI capabilities without incurring training costs and threatens US national security.

The White House is asking OpenAI to slow roll the release of its new model over safety concerns

The White House has asked OpenAI to delay the full release of GPT-5.6, initially limiting access to government-approved partners to address national security concerns.

The AI Incident Reporting Act would require AI companies to report critical safety incidents within seven days, with Congress notified within 48 hours for the most serious events.

UAE Establishes Federal Artificial Intelligence and Data Authority

The UAE created a new federal authority to govern AI, data, and digital government, consolidating oversight under one body.

Australia Launches AI Accelerator Round to Boost Industry Commercialisation

The Australian government announced a $50 million AI-focused Cooperative Research Centre funding round to accelerate AI commercialization across sectors.

Aidoc Receives FDA Breakthrough Device Designation for AI Radiology Report Drafter

Aidoc's First Read AI, which drafts radiology reports, received FDA Breakthrough Device Designation, speeding its path to market.

Industry & market moves

SiMa.ai Partners with Mistral for Drone Intelligence

SiMa.ai and Mistral Solutions have partnered to create a validated drone platform combining SiMa.ai's hardware with Mistral's drone expertise for autonomous drone intelligence.

Former Anthropic Researchers Launch Mirendil at $1 Billion Valuation With $200M Seed Round

A startup founded by ex-Anthropic researchers raised a $200 million seed round at a $1 billion valuation from top-tier investors.

China’s DeepSeek plans to double its workforce after $7.4 billion fundraising

DeepSeek, the Chinese AI firm, plans to double its workforce to accelerate AGI development following a $7.4 billion funding round.

Patronus AI lands $50M to build ‘digital worlds’ that stress-test AI agents

Patronus AI raised a $50M Series B to create digital world models for testing AI agents, expanding its research and commercial efforts.

Permira and Warburg Pincus complete Clearwater Analytics acquisition

An investor group led by Permira and Warburg Pincus completed the buyout of Clearwater Analytics, taking it private to accelerate its AI-driven roadmap.

TensorX and Solstice: $1 billion for dominant AI in Europe

TensorX and Solstice announced a $1 billion financing facility to build AI hardware and data centers in Europe, aiming for onshore AI sovereignty.

Pro Medicus Echo IQ Deal Expands AI Healthcare in Australia

Pro Medicus signed a financing and reseller deal with Echo IQ to integrate AI algorithms for cardiology into its Visage 7 platform globally.

AI product & feature launches

Convr Makes Commercial P&C Risk Context Engine Available to AI Agents via MCP

Convr released its Risk Context Engine for MCP, allowing any MCP-compatible AI agent to access commercial insurance underwriting risk data.

IBM Debuts World's First Sub-1 Nanometer Chip Technology

IBM unveiled a 0.7nm chip architecture using 3D nanostack technology, doubling transistor density over 2nm chips.

Research with immediate practical relevance

No significant new developments.