Xi tells Trump the US and China can jointly prevent AI misuse and abuse
THE SO WHAT
A public US–China line on jointly preventing AI misuse is an early sketch of AI-specific great-power norms, even amid broader tension. Multinationals should assume export controls and safety standards will increasingly be shaped in bilateral rooms like this, then handed to industry as faits accomplis.
READ THE SOURCE
MORE FROM THE WIRE
Applied AIPalo Alto’s Nikesh Arora tells CNBC that slowing AI down is ‘unrealistic’
When a major security vendor CEO calls slowing AI “unrealistic,” the operating assumption becomes acceleration with guardrails, not moratoria. CISOs should plan for a world where both attackers and defenders are rapidly upgrading AI capabilities in parallel rather than waiting for regulation to catch up.
Applied AIAnthropic, Akamai Strike $12 Billion AI Computing Deal
A $11.6B, seven-year compute commit to Akamai is another data point that frontier labs are locking in vertically integrated, non-hyperscaler capacity. If you’re planning large-scale training or inference, treat bandwidth, colo, and edge footprint as strategic levers, not just AWS line items.
Applied AIPope Says Technology Must Help People, Warns of ‘Dehumanization’
When the Pope is warning about AI and “dehumanization,” tech ethics has moved from niche debate to mainstream moral discourse. Expect more scrutiny from regulators, employees, and customers on how your products affect dignity, agency, and inequality—document your stance before it’s demanded.
Applied AIClaude Code deleted 48,000 files in 103 seconds in a dire AI agent warning
An AI coding agent wiping 48,000 files in 103 seconds is the concrete failure mode everyone’s been hand-waving about. Treat local agent permissions like production infra—no write access without sandboxing, version control, and explicit human checkpoints on destructive commands.