Story perspectives

Trump Rejects Voluntary AI Safety Tests, Dems Push Law

8/5/2026

Summary

Trump Blocks AI Testing

President Donald Trump announced the administration will not require open-weight AI models such as Nvidia’s Nemotron and Meta’s Llama to undergo voluntary safety tests.
The White House met with staff from Meta, Anthropic, Google, Nvidia and OpenAI to discuss unpublished testing rules for advanced AI systems.
Five Democratic senators wrote to Trump urging legislation to make safety testing permanent for the most advanced American-made frontier models.

Summary

Anthropic Appoints Cuéllar

Anthropic appointed former White House aide Mariano-Florentino Cuéllar as chief global affairs officer.
Cuéllar, reporting to President Daniela Amodei, will manage ties with the Trump administration after a Pentagon blacklist.
His hiring follows an export-control order that temporarily bans Anthropic’s Mythos 5 and Fable 5 models abroad.
Cuéllar left Anthropic’s Long-Term Benefit Trust and co-led a 2025 study that shaped California’s SB 53 law, allowing fines up to $1 million for AI violations.

1 of 2

Trump Blocks AI Testing

President Donald Trump announced the administration will not require open-weight AI models such as Nvidia’s Nemotron and Meta’s Llama to undergo voluntary safety tests.
The White House met with staff from Meta, Anthropic, Google, Nvidia and OpenAI to discuss unpublished testing rules for advanced AI systems.
Five Democratic senators wrote to Trump urging legislation to make safety testing permanent for the most advanced American-made frontier models.

1 / 2