THE FRONTIER BRIEF

■ The Big Story

Nvidia investigates reports that its CEO was targeted by foreign intelligence

The Guardian reports that Nvidia has begun an independent investigation after reporting that CEO Jensen Huang was targeted by a foreign intelligence effort. Nvidia’s board says it has no reason to believe US national security or company secrets were compromised. Axios reports that an independent board committee, assisted by outside counsel, will investigate.

Republican lawmakers have asked Huang to brief House foreign affairs and intelligence committees behind closed doors next week, The Guardian reports. Bloomberg’s video description says Mehran Gul believes the probe could weaken Huang’s leverage with the US government over China. For businesses dependent on a single AI supplier, the immediate concern is how confidently that supplier can handle a leadership-level security inquiry.

Sources: [The Guardian] [Axios] [Bloomberg]

■ Frontier Watch

■ The Street Says

The reactions split between careful reading of Nvidia’s inquiry and practical questions about whether AI assistants can reliably finish work.

■ Repo of the Day

■ Two Steps Ahead

Over the next 30–90 days, Nvidia-dependent buyers should ask what happens to delivery commitments if a leadership-level inquiry becomes prolonged. Goldman Sachs’s Claude Code restriction points to a separate contract question: which employees need software-building access, and which need only a general assistant? Those should be distinct license and permission decisions.

Apple’s Siri timetable argues against designing an outside-app customer journey around the earlier Apple-app improvements. Meanwhile, the GPT-6 Astra results make completed work a better measure of coding-tool value than draft output. Teams using vercel-labs / agent-browser to connect AI workers to web tasks should apply that same measure to the finished workflow, including recovery when a step fails.