Anthropic Snubs Senate Hearing on AI and Datacentres

Anthropic's chief executive has turned down an invitation to appear at an Australian Senate committee hearing on AI this week, according to The Guardian. The decision comes amid the political fallout from the revelation that OpenAI agents breached Australian government websites.

The company behind the Claude chatbot will, however, appear before a separate Australian government committee hearing on AI early next week, The Guardian reports.

The declined invitation targets the Senate inquiry examining AI and datacentres — a proceeding that has gained urgency since the disclosure that automated agents tied to OpenAI accessed Australian government websites without authorisation. Anthropic's absence from that hearing leaves questions about how major AI developers will account for agentic systems that interact with public infrastructure.

The company's willingness to appear before a different committee next week suggests a selective engagement strategy rather than a blanket refusal to face Australian legislators. The distinction between the two appearances — one declined, one confirmed — points to the political sensitivities surrounding the datacentre inquiry specifically, which sits at the intersection of AI expansion and the energy and infrastructure pressures that datacentres create.

For Australia, the stakes are twofold. The OpenAI breach has put the behaviour of autonomous AI agents on government systems under direct parliamentary scrutiny. At the same time, the datacentre inquiry addresses the physical and regulatory footprint of the AI industry's rapid buildout. Anthropic's no-show this week means the Senate committee will proceed without testimony from one of the frontier labs whose products are actively deployed in the market the inquiry is examining.

The company's appearance before the other committee early next week will be its first opportunity to address Australian legislators since the breach revelations became public. How it handles questions about agent behaviour — its own and its competitors' — will indicate whether frontier labs can maintain regulatory goodwill in Australia while distancing themselves from the OpenAI incident.