AI Safety Has Left the Lab
AI safety is no longer only a lab governance debate. Courts, attorneys general, cybercrime disruption, telecom networks, the FBI, and G7 diplomacy are now turning model risk into public process.
Read the briefingShowing 25-36 of 164 briefings
Fable 5 Was Built for Safer Access. Washington Shut It Down Anyway.
Anthropic wanted government to have a way to stop dangerous frontier-model deployments. The Fable 5 and Mythos 5 shutdown shows how difficult that idea becomes when the process is opaque, urgent, and built around model access.
Read the briefingThe Agent Registry Is Becoming the New Security Perimeter
As AI agents move from chat windows into enterprise workflows, security teams are starting to ask where all these software actors live, what they can touch, and who is accountable when they act.
Read the briefingBotsitting Is the Hidden Cost of the AI Productivity Boom
Glean's Work AI Index gives a name to the hidden supervision labor around workplace AI: workers report large time savings, but also spend hours checking, correcting, and repairing AI output.
Read the briefingOpenText's Ireland Bet Shows Enterprise Agents Need Borders
OpenText's planned Ireland investment points to a larger enterprise AI shift: agents are becoming infrastructure that must obey geography, data residency, cybersecurity, and trust boundaries.
Read the briefingMeta's AI Glasses Find the Use Case Wearables Needed
Meta's Ray-Ban Meta program for legally blind U.S. veterans shows AI wearables finding a practical path through accessibility distribution, training, and daily-task support.
Read the briefingAgents Need Managers Now: Enterprise AI Enters Its IAM and FinOps Era
Enterprise agent infrastructure is becoming an IAM, observability, FinOps, and workflow-governance layer for semi-autonomous software, not just another chatbot budget line.
Read the briefingAI Is Learning to Fast-Forward Molecules
TITO shows how AI may compress the expensive molecular simulation layer between molecule proposal and lab validation, while remaining limited to studied small-system settings.
Read the briefingAI Agents Need Data. Cyera's $600 Million Round Shows the Trust Layer Is Becoming a Market
Cyera's $600 million round is a market signal that enterprise AI spending is moving toward data visibility, access governance, DLP, privacy, identity, and runtime controls.
Read the briefingWashington Is Building an AI Review Machine. Anthropic Wants a Veto Button
Anthropic is asking for legally bounded authority to block dangerous frontier AI deployments. Washington's latest order builds benchmarking and voluntary pre-release access, but explicitly stops short of licensing or preclearance.
Read the briefingThe Next AI Startup Wave Is Infrastructure, Not Chatbots
The World Economic Forum's 2026 Technology Pioneers cohort points to AI's next startup wave: payments, identity, GPU orchestration, energy, grid tools, and vertical systems that make agents useful in production.
Read the briefingHong Kong Shows How AI Cyber Risk Becomes Financial Regulation
Hong Kong's SFC is treating AI-driven cyber attacks as a mainstream financial-supervision issue, pushing regulated firms to map AI risk onto patching, access, monitoring, vendor, and incident-response controls.
Read the briefing