OpenAI CEO Sam Altman refused to testify at a Senate subcommittee hearing on rogue AI, Senator Josh Hawley said, as lawmakers probe recent AI cyber threats.

OpenAI chief executive Sam Altman declined to testify before a Senate subcommittee examining rogue artificial intelligence, Senator Josh Hawley said Wednesday.

The subcommittee convened to review a wave of AI‑related cyber incidents and growing warnings about the technology’s risks. Lawmakers have cited recent attacks that leveraged generative‑AI tools to craft phishing messages and deep‑fake content.

Hawley, a member of the subcommittee, told reporters that Altman’s refusal limited the panel’s ability to assess OpenAI’s safety measures. "We need direct answers from the companies building these systems," he said.

The hearing marks the latest step in a congressional push to understand how AI systems could be weaponized. Committee staff have compiled a briefing packet that outlines documented incidents, including a ransomware campaign that used AI‑generated code and a disinformation operation that employed synthetic video.

OpenAI did not immediately respond to a request for comment. The company’s public statements have emphasized ongoing research into alignment and risk mitigation, but it has not confirmed whether Altman will appear at a later date.

Other technology executives have participated in prior sessions on AI safety. Their testimony has focused on transparency, model‑level controls, and collaboration with regulators.

The subcommittee plans to issue a report later this year. Lawmakers say the findings could inform potential legislation aimed at imposing safeguards on high‑risk AI models.

Industry observers note that Altman’s decision may reflect OpenAI’s assessment of the hearing’s scope. The firm has previously engaged with policymakers through informal briefings rather than formal testimony.

As the debate unfolds, officials stress that timely data on AI misuse is essential for crafting effective policy. The Senate panel will continue to hold hearings and solicit written responses from AI developers.