Anthropic launches critical infrastructure AI program, free OSS scanner

Anthropic has launched a program to deploy its frontier AI models and on-site engineers to security teams protecting power grids, water systems, and other critical infrastructure. Alongside it, the company released OSS Scanner, a free tool that provides open-source projects with periodic vulnerability scans from its most advanced models. Both initiatives fall under Anthropic’s new Cyber program.

The timing is notable. When we covered AWS open-sourcing a 2-billion-parameter decision model on 1 October, it marked a growing trend of cloud providers releasing lightweight AI tools. Anthropic’s offering, however, targets a different audience: the maintainers of open-source software that underpins critical infrastructure. The free scans could reduce the barrier for smaller projects that lack dedicated security teams, though the company has not disclosed how often scans will run or what criteria determine eligibility.

Anthropic’s decision to embed engineers with infrastructure security firms suggests a hands-on approach to AI safety, aligning with its stated mission of building reliable systems. It also raises questions about how these models will be used in practice. Critical infrastructure operators have historically been cautious about adopting AI, particularly from vendors without a long track record in the sector. The program’s success may hinge on whether Anthropic can demonstrate tangible improvements in threat detection without introducing new risks.

For now, the company has not shared details on which infrastructure operators are participating or how many open-source projects have signed up for OSS Scanner. Observers will be watching to see if other AI labs follow suit—or if Anthropic’s model becomes a template for how closed-source AI can engage with open ecosystems.

Sources: siliconangle.com