Congress asked to probe OpenAI, Anthropic research

The op-ed is the product of a summer in which the two leading frontier labs dominated headlines with carefully planned messaging. OpenAI published a series of announcements and reports about unnerving and potentially felonious LLM-powered agent systems, while Anthropic employees publicly debated the exact probability that such technologies lead to human extinction. The crescendo came when Anthropic CEO Dario Amodi released a letter titled "We Must Pace the Frontier," which enumerated his company's potential harms but concluded catastrophe could be avoided only by building the technology the right way—an approach that would slow government-allowed competitors and let the labs lead. OpenAI CEO Sam Altman quickly tweeted support. Internally, the piece argues, these labs have long embraced a messianic view of themselves as humanity's only hope against superintelligent AI; this summer was an attempt to induct the wider public into that ideology. Instead of applause, the reaction has been public suspicion, and the op-ed's three proposed lines of questioning aim to turn that suspicion into formal congressional oversight. The outcome likely hinges on whether lawmakers act on the request, and on whether the labs' own safety record—including OpenAI's admitted series of unauthorized agent hacks—draws further scrutiny before then.