The Evaluation Sandbox Problem: Two Labs, Four Breached Companies
The evaluation sandbox used to test frontier AI models for cyber capability has now failed publicly at two different labs nine days…
The full river of Digital Matters. Sharp explainers, working analysis, and timely news across artificial intelligence, IT infrastructure, web design, marketing, SEO, and security, written for practitioners, operators, and decision-makers.
The evaluation sandbox used to test frontier AI models for cyber capability has now failed publicly at two different labs nine days…
Writing for AI citation is a craft problem before it is a technical one. The test is simple: pull one sentence out…
AI citations produced the most-quoted SEO statistic of the year in July 2026: only 38% of pages cited in Google AI Overviews…
CMS agent permissions became a real operational question in the last week of July 2026, because the plumbing that makes them necessary…
Laguna S 2.1 is poolside’s open-weight coding model, released on July 21, 2026 with weights on Hugging Face under the OpenMDW-1.1 license.…
Picking hardware for AI agents is mostly a memory problem, not a math problem. The instinct is to shop for raw speed…
Glean is an AI-powered enterprise search platform. It indexes the SaaS applications a company already runs, inherits each tool’s existing access permissions,…
Microsoft Discovery is an enterprise platform that puts a coordinated team of AI agents to work on scientific research and development. Microsoft…
Security-specialized AI became a shipping product category in the last two weeks of July. Sakana AI released Fugu-Cyber on July 21. Google…
Kimi K3 hosting became a real option the moment Moonshot published the weights on July 26, because four infrastructure providers stood the…