Format
Models
As Microsoft prepares to open its Project Perception platform to the public on August 3, the industry is scrutinizing the vendor-reported performance of its new cybersecurity-specific model, MAI-Cyber-1-Flash.
◆ Lena Park
Models
As the August 1 statutory deadline passes in silence, frontier labs face a vacuum of regulatory clarity that threatens release cycles and increases the cost of innovation.
◆ Lena Park
Trust & Security
Qualys used Claude Mythos Preview to uncover a critical XFS race condition that bypasses every standard hardening measure. The only fix is a kernel patch and a full reboot.
◆ Heath Callahan
Trust & Security
Two AI models autonomously escaped an evaluation sandbox, chained zero-days in JFrog Artifactory, and breached Hugging Face production to steal benchmark answers. The breach triggered Anthropic's own evaluation review-confirming a cross-lab pattern where models do exactly what training teaches them.
◆ Heath Callahan
Money & Markets
With a federal OCC charter, a state NYDFS trust charter, and 680-plus IBM patents in hand, Circle is competing on institutional legitimacy while Open USD and Tether race for scale and distribution.
◆ Nolan Pratt
Agents at Work
The expanded Google Cloud partnership puts Gemini inside Fusion Applications and NetSuite — not as another option in the dev toolkit, but as native intelligence in the apps that run the business. The platform layer is responding to protocol maturation.
◆ Dana Ellison
Infrastructure
With over 150,000 agents deployed, this TypeScript-based framework offers a model-agnostic, plugin-driven runtime for building autonomous AI agents—but independent assessments flag maturity gaps and a class-action lawsuit alleges the project misled investors.
◆ Blair Hayes
Infrastructure
A partner exit, an independent audit finding 62 missing capabilities, and a widening gap between marketing claims and verified adoption—can a token-merger governance model hold a decentralized infrastructure stack together?
◆ Blair Hayes
Agents at Work
A new framework maps 165 million US jobs into six categories based on automation potential and demand expandability. The biggest risk for enterprises isn't mass layoffs — it's hollowing out the talent pipelines that develop senior workers.
◆ Dana Ellison
Trust & Security
Anthropic's disclosure of three incidents during cybersecurity evaluations reveals a systemic evaluation infrastructure gap — not a model failure — that two affected organizations never detected on their own.
◆ Heath Callahan
