NewsNTech
The speed of capability gains in large AI systems has consistently outpaced the development of tools to verify those systems behave as intended. That constraint is now a political contest as well as a technical one.
More than a dozen researchers inside OpenAI and Anthropic have called for the industry to slow down; Trump has publicly dismissed warnings that artificial intelligence poses an extinction-level threat to humanity.
The researchers raising those alarms work inside OpenAI and Anthropic.
Their position carries different weight than outside commentary: they have direct visibility into the systems in development and have concluded that rapid advances warrant a pause.
Keep reading