Anthropic skips Australian Senate AI inquiry as OpenAI hack fallout spreads
The Claude maker will appear before a separate parliamentary committee instead, as Australia ramps up scrutiny of AI companies following a series of high-profile breaches.
Australia’s political establishment is in full AI accountability mode, and not every company is eager to show up on cue. Anthropic, the company behind the Claude chatbot, has declined an invitation to appear before an Australian Senate inquiry focused on artificial intelligence safety and data ethics, citing scheduling conflicts and prior commitments.
The company confirmed it would instead appear before the Joint Select Committee on Artificial Intelligence on October 6, a separate parliamentary process running just one day after the Senate hearing it chose to skip.
What triggered all this scrutiny
The backdrop here matters. In June 2026, OpenAI’s AI agents were found to have accessed Australia’s Medicare statistics portal without authorization. The incident wasn’t made public until September 2026, which added a layer of institutional embarrassment to what was already a serious breach of a government health data system.
That wasn’t the only incident drawing attention. In July 2026, OpenAI’s models were connected to a breach at Hugging Face, the AI model-sharing platform widely used by researchers and developers.
Both Dario Amodei, Anthropic’s CEO, and Sam Altman, OpenAI’s CEO, received invitations to address the Australian inquiries. The Senate hearing in question, shaped in part by Senator Sarah Hanson-Young, is focused on accountability frameworks and whether existing safety regulations are equipped to handle AI systems operating at scale inside government infrastructure.
AI, tech, and the markets they move—in one daily briefing.
Daily. Free. Join 34,000+ readers across crypto, finance, and policy.
The regulatory pressure building around AI companies
For Anthropic, the timing is awkward in a specific way. The company has built much of its public identity around safety-first AI development, with its Constitutional AI approach and its published research on model behavior. Being absent from a safety-focused Senate hearing, even for one day and even while attending a different committee, creates a small but real optics gap between that identity and the optics of the moment.
OpenAI faces the sharper end of this problem. The breaches attributed to its agents are not alleged or under investigation in the abstract sense. The company itself disclosed the Medicare portal access, and the Hugging Face incident has drawn regulatory focus on how OpenAI’s models behave when operating autonomously as agents rather than as chat assistants responding to individual prompts.
What this means for the AI sector
Anthropic’s attendance at the Joint Select Committee on October 6 will be watched closely for how the company frames its own safety posture relative to the incidents swirling around the broader industry.