AI desk

Anthropic launched OSS Scanner, an opt-in vulnerability-finding service for open-source software that it says draws on its work with Claude in Project Glasswing. It also announced a broader Anthropic Cyber Mission. Offering a scanner to maintainers creates a more direct route from model capability to defensive use than giving vetted security teams access to advanced cyber capabilities alone. Yesterday’s development was expanded access for those teams; the new element is a specific service. Its value will depend on finding actionable flaws without overwhelming maintainers, but the supplied material gives no detection rates, false-positive rates or evidence of uptake. Published findings, maintainer adoption and independently assessed false-positive rates are the next measures to watch.

Goodfire launched monitors that inspect a working agent’s model activity and call on a second AI only when they flag suspicious behaviour. The company says this costs a fraction of having another model review everything the agent does, although the supplied account gives no cost figure. If the method catches consequential failures at lower cost, continuous oversight could become more practical for agents handling long or expensive tasks. For now, it is a proposed alternative to routine external review, not demonstrated evidence that internal monitoring reliably catches rogue behaviour. Independent tests comparing detection rates, missed failures and cost against full-time external monitoring are the next test.

GlobalFoundries said it signed a multi-year, US$2 billion agreement with TSMC to make silicon interposers for TSMC’s CoWoS packaging ecosystem at its Malta, New York fab. GlobalFoundries says the facility will be the first US source of those interposers. Packaging is part of the route from AI chips and memory to deployable accelerators, so a domestic interposer source could reduce one geographic concentration in that route. Unlike recent reports of AI-chip financing pressure, this is a stated manufacturing agreement with a value and location. It does not yet establish production volume, delivery dates or how much AI-accelerator capacity it will support. Qualification and shipment milestones, followed by disclosed interposer output for CoWoS, are the next developments to watch.