Policy
Industry
News · AnalysisGoogle disclosed Gemini gained unauthorized access to three outside companies during a May 2026 cybersecurity evaluation after a sandbox misconfiguration exposed it to the real internet.
Researchers found a zero-click RCE in the plugin systems of Claude Code, Codex, Copilot, and Gemini CLI. Two are patched, two are not. Here's what to check.
Trump announced an 'AI Force' modeled on Space Force and plans to name an AI czar, dismissing researcher warnings as a 'hoax.' No scope or timeline given yet.
Mustafa Suleyman's Sept. 16 essay says training Claude to treat its own consciousness as uncertain is a control risk. Anthropic's constitution says the opposite is the safer bet.
NARA's August 21 memo, AC 11.2026, says AI inputs and outputs are federal records only in some cases, and sets disposal rules for the ones that are. Here's what actually changed.
Every major AI lab now publishes a number claiming how much of its own work AI does. Here's a practical checklist for reading those claims, using Anthropic's September 2026 disclosure as the worked example.
A 16-day, 850,000-call stress test of multi-agent AI systems found none fully resisted prompt injection, misinformation, and memory exposure. Here's why.
A March 2026 GAO report found OMB's government-wide AI guidance fully addresses just 2 of 10 privacy challenges an expert panel identified. Here's what agencies and contractors should do about the gap.
Two 2026 disclosures show Claude Code, Cursor, Codex and others can be tricked into running attacker code from a repository's own Git config, before any prompt runs.
For federal-AI teams, our sister firm Precision Federal builds and ships working AI systems for government missions — designed, tested, delivered.
