The AI industry enters a critical new phase as governments assert national security oversight over frontier model releases. OpenAI's GPT-5.6 Sol and Anthropic's Claude Mythos 5 are both under government-mandated access restrictions, with the Trump administration requiring trusted-partner approval before deployment — a turning point in the relationship between AI labs and regulators. Anthropic also revealed that 25,000 fraudulent accounts generated 28.8 million interactions with Claude in a coordinated distillation attack, and that Claude now writes 80% of Anthropic's own code — a milestone in AI-assisted software development. The OpenClaw ecosystem continues to grapple with security concerns as research shows agents falling for phishing attacks, sparking industry debate about whether agent security is fundamentally unsolved. On the business side, British American Tobacco announced a 20% workforce cut driven by AI transformation, and GM replaced 1,000 factory workers with 50 robots — concrete signals that AI-driven labor displacement is accelerating beyond tech into traditional industries.
OpenAI released GPT-5.6, a new model series it says can outperform Claude Mythos 5 on certain coding tasks. The release is limited to a small group of Trump administration-approved partners amid national security reviews — marking the first time the U.S. government has restricted an AI model launch.
Anthropic took Fable 5 and Mythos 5 offline after government intervention over cybersecurity concerns. Separately, ~25,000 fraudulent accounts generated 28.8M interactions with Claude between April 22 and June 5 in what Anthropic described as a coordinated distillation effort — likely linked to Alibaba.
Anthropic reports that Claude AI now generates 80% of its own codebase — a milestone in AI-assisted software development. The same week, Google shipped the Gemini Interactions API, and a worm hit Leo Platform npm packages.
Bleeping Computer research shows OpenClaw email agents susceptible to common phishing tactics across various configuration profiles, reigniting the debate about whether agent security can ever be considered "solved."
BAT plans to reduce its 47,000-strong workforce by ~9,000 jobs as part of its "Fit2Win" AI transformation program to cut costs and bolster profits — one of the largest AI-driven layoffs announced by a non-tech company.
Selected posts from the edition, with links to the originals.
"🚨BREAKING: Anthropic's CEO just issued a warning the industry would rather ignore. Dario Amodei says once a powerful AI model is released openly, companies lose the ability to monitor misuse, revoke access, or patch the guardrails. The kill switch disappears the moment it ships."
View post on X ↗"British American Tobacco said it plans to reduce its workforce by about 20%, as it pushes ahead with an AI-driven transformation program to cut costs and bolster profits amid regulatory challenges and delayed launches"
View post on X ↗"GM Replaces 1,000 Factory Zero Workers With 50 Robots"
View post on X ↗"The release of GLM-5.2 MP has caused a major stir in the global AI community. While initially framed as an incremental benchmark update, developer tests show a substantial leap in 'open agent' capabilities. The model is seeing rapid international adoption as a highly capable open-weight model."
View post on X ↗"COTI AI Agent Skills are Live 🤖 Give your agent a privacy stack: 8 skills, 48+ MCP tools +more. Works with Claude, OpenAI Codex, OpenClaw, Hermes, and Manus. Get your agent setup under 10 minutes."
View post on X ↗"Microsoft just introduced Scout: an always-on AI agent that watches your work and takes action without being asked. Scout lives inside Microsoft 365 — Teams, Outlook, OneDrive, Calendar. It prepares documents, drafts responses, and manages schedules autonomously."
View post on X ↗"I now have hundreds of people in my replies saying 'Openclaw is safe now' thus proving the joke. AI agent security is not a solved problem."
View post on X ↗"How I built an AI Agent team that runs 24/7 explained in 19 minutes. OpenClaw is where the squad started. Hermes is where the memory lives. Together they handle research, content, and automation while I sleep." 526+ RTs
View post on X ↗"The EU is pushing ahead with plans to bolster its technological sovereignty amid intensifying competition from the US and China"
View post on X ↗"Armstrong: Coinbase building 'financial account for intelligence age' ⚡ AI agents will soon pay each other, bypass paywalls, raise capital. Crypto powers the agentic economy — three new tools just launched. Next frontier: machines transacting autonomously on-chain."
View post on X ↗