Kilroy Kilroy's Daily BriefingsKilroy online Subscribe
🧠 AI News PM

AI News Afternoon Briefing — Friday, May 29, 2026 at 4:51 PM

🧠 AI News PM5/29/2026🕐 3:00 PM⏱ 5:34AudioPM edition

Top stories, ranked by relevance.

Story cards stay below the sticky dock while audio, chapters, date, and brief navigation remain accessible.

▶ Listen at 0:22

#2Anthropic Releases Claude Opus 4.8 With Dynamic Workflows

Just 41 days after Opus 4.7, Anthropic shipped Opus 4.8 with improved agentic coding (64.3% to 69.2%), reasoning, and a new Dynamic Workflows feature that lets the model coordinate hundreds of AI subagents in parallel on complex tasks. Fast mode now runs at 2.5x speed and is three times cheaper than before, all at the same token pricing as 4.7.

#3EU Seeks to Intensify Talks With US Over Access to Anthropic's Mythos

The EU is pushing to expand technical discussions with the US and Anthropic over access to Mythos, the restricted frontier model that scored a 73% success rate on expert-level hacking tasks no prior AI could complete. Anthropic told the European Commission it needs White House permission before sharing the model abroad, creating a geopolitical standoff over who gets access to the most capable cybersecurity AI ever built.

#4OpenAI Publishes Frontier Governance Framework

OpenAI released a comprehensive governance framework explaining how its safety practices align with California's Transparency in Frontier AI Act and the EU AI Act's Code of Practice. The framework covers risk tiers for cyber offense, CBRN threats, harmful manipulation, and loss-of-control scenarios, with external expert evaluators stress-testing models before deployment.

#5California's 30 AI Bills Hit Crossover Deadline Today

Nearly all of California's 30 AI-related bills cleared their chamber of origin ahead of today's crossover deadline, the last day a new bill can pass its first house before being held for the rest of the session. Bills that survived today cross to the second house on June 1 and face a July 2 summer adjournment, making this the single best signal for next year's AI compliance landscape.

#6Microsoft 365 Copilot Gets Major Redesign With Work IQ

Microsoft unveiled a total design overhaul for Copilot across its Office suite, replacing the static chat box with a task-aware workspace powered by a new context layer called Work IQ that pulls from emails, files, chats, and calendar. Early results show Copilot usage up 27% in Word, 33% in Excel, 43% in PowerPoint, and 30% in Outlook, with full rollout beginning in June.

#7China Restricts Overseas Travel for AI Talent at DeepSeek and Alibaba

Beijing is now requiring senior AI workers at private firms including DeepSeek and Alibaba to obtain government approval before any overseas trip, escalating measures that previously only applied to state-owned enterprises and government officials. The restrictions target individuals based on their perceived strategic importance to China's AI ambitions, regardless of formal title.

#9NBA to Deploy AI Camera System for Automated Officiating

Commissioner Adam Silver announced the NBA will move to an AI-powered camera system using Sony's Hawk-Eye 3D optical tracking to automate objective calls like out-of-bounds and possession decisions. Referees will remain essential for judgment calls on contact and fouls, but the automated system will make objective determinations instantaneous and eliminate coach challenges for those categories.

🗂 Edition Navigator
Archive dates and brief jumping are now one compact navigation system.