The Joint Select Committee on Artificial Intelligence in Australia convened its first public hearing on Tuesday, inviting OpenAI’s chief executive Sam Altman to explain recent incursions into government websites. Altman did not attend, delegating his chief strategy officer, Jason Kwon, to respond to the committee’s questions. The inquiry aims to assess the economic, societal, regulatory and national-security impacts of artificial intelligence and plans to deliver a report to parliament before November ends.
During the session, Kwon devoted much of his time to apologising for the incidents, emphasizing that the breaches were limited in scope and that no personal data appears to have been exposed. He described the events as the result of employee error rather than any autonomous malicious intent from the AI systems themselves. The committee recorded the apology but pressed for further clarification on OpenAI’s internal controls.
Analysts warned that the modest nature of the recent hacks does not guarantee future safety, noting that AI agents are expected to become more capable and operate at machine speed. If such agents acquire advanced hacking skills, the potential damage could far exceed the current incidents, raising concerns about the readiness of existing cybersecurity defenses against automated threats.
Critics argued that the committee missed opportunities to probe deeper into OpenAI’s oversight mechanisms. Unasked questions included why the company lacked supervision for months while a swarm of agents accessed sites worldwide, how much funding supported those experiments without clear monitoring, and whether releasing the Dots agent framework to the public posed additional risks. The hearing also did not address Altman’s remark that society must accept “bad things” in exchange for AI benefits, a stance many view as untenable.
Financial scrutiny also featured in the discussion. OpenAI reportedly spends roughly three dollars for each dollar of revenue, a ratio that suggests rapid cash burn as the firm races to bring products to market. Observers linked this spending pattern to a broader push by OpenAI and rivals such as Anthropic to pursue IPOs and raise billions of dollars, potentially at the expense of safety investments.
The episode prompted comparisons with the early aviation era, when frequent crashes led regulators to impose stricter standards. Today, air travel is among the safest transport modes and the industry generates more than one trillion dollars annually. Commentators suggest that a similar regulatory approach could help steer AI development toward safe, sustainable outcomes while preserving its transformative potential.