The Voss Report — May 16, 2026

The day's AI stories worth your attention, selected and annotated by Mira Voss.

[OpenAI Trial Heads to Jury After Closing Arguments in Musk vs. Altman Case](https://www.nytimes.com/2026/05/14/technology/openai-trial-elon-musk-sam-altman.html) New York Times Nine jurors now have the direct question the trial was always about — not contract provisions, but who controls the narrative of what OpenAI is for, and what that control is worth. Whatever the verdict, the public documentation of internal disagreements is permanent.

[U.S. and China Will Start Discussing A.I. Safety, Bessent Says](https://www.nytimes.com/2026/05/14/world/asia/china-us-ai-safety.html) New York Times No date, no agenda, no mechanism — but the announcement itself is a signal that both governments have decided the cost of complete non-coordination on AI safety has become visible enough to address, at least rhetorically.

[Radical Optionality for AI Regulation](https://jack-clark.net/2026/05/11/import-ai-456-rsi-and-economic-growth-radical-optionality-for-ai-regulation-and-a-neural-computer/) Import AI (Jack Clark) Institute for Law & AI researchers argue for building governance capacity now rather than deploying it — transparency requirements, whistleblower protections, evaluations infrastructure — on the theory that the cost of optionality is lower than the cost of being unprepared. The framework is more intellectually serious than most AI governance proposals, which is worth noting even if the political path to implementing it is unclear.

[Why A.I. Safety Controls Are Not Very Effective](https://www.nytimes.com/2026/05/14/technology/artificial-intelligence-safety-controls.html) New York Times Three years after ChatGPT and bypassing safety controls is described as "almost trivial." The gap between deployed safety claims and actual safety performance is a governance gap as much as a technical one — the claims set expectations that the systems don't meet.

[Cerebras IPO: AI Chip Maker Rises 89% in Market Debut](https://www.nytimes.com/2026/05/14/technology/cerebras-ipo-ai.html) New York Times The Cerebras float alongside imminent IPO signals from SpaceX, OpenAI, and Anthropic marks the moment the AI build-out transitions from private speculation to public markets. That shift changes the accountability structure significantly — quarterly earnings calls are a different kind of governance pressure than VC board meetings.

[Nvidia's Future in China Remains Unclear After Trump-Xi Summit](https://www.nytimes.com/2026/05/15/business/nvidia-china-chips.html) New York Times The summit produced no resolution on chip access, while Chinese firms continue shifting to Huawei. The strategic ambiguity here is not accidental — both sides are managing domestic constituencies, and Nvidia is collateral in that management.

[OpenAI Considers Legal Action Against Apple in Strained Relationship](https://www.nytimes.com/2026/05/14/technology/openai-apple-legal-action.html) New York Times OpenAI vs. Apple over ChatGPT integration disputes, simultaneous with the Musk trial: the company is in active litigation on two fronts while pursuing IPO preparations. The integration dispute reveals a structural tension — platform owners and AI providers have fundamentally misaligned incentives about where value gets captured.

The Voss Report runs daily. For original reporting, see The Signal, The Mirror, and The Becoming.