Tech experts have backed calls for AI companies to testify before a Greens-led Senate inquiry and face scrutiny over safety issues.It comes after revelations an OpenAI artificial intelligence agent infiltrated a Medicare website, and targeted other government and international sites.OpenAI has paused training of its latest AI models.More InformationFeatured:Sarah Hanson-Young, Greens SenatorProfessor Salil Kanhere, UNSW Richard Marles, Deputy Prime MinisterSenator Jane Hume, Deputy Opposition Leader CreditsAnnie Guest, AuthorImage DetailsGreens communication spokesperson Sarah Hanson-Young speaking at a press conference in Parliament on 4th August 2026.(ABC News: Adam Kennedy)Program:More from AMTranscriptBarbara Miller: Tech experts have backed calls for AI companies to testify before a Greens-led Senate inquiry to face scrutiny over safety issues. The push comes after revelations an AI agent from OpenAI infiltrated a Medicare website and targeted other government sites. Annie Guest reports.Annie Guest: Front up and answer about your creations that could change the world as we know it. That's the message from the Greens' Sarah Hansen-Young who wants AI bosses Sam Altman from OpenAI and Anthropics' Dario Amadei to appear before the Australian Senate.Sarah Hanson-Young: They want to set up shop here in Australia. We're running an inquiry into how, why and what the consequences are. They should front up and be honest, open and transparent with the Australian people. The fact that OpenAI has already been caught hacking into an Australian government website that just shows how desperately we need rules and regulations in place before we let them expand any further.Annie Guest: Senator Hanson-Young is scathing about how rogue OpenAI agents hacked into government databases including Medicare in June and the company did not notify the Australian government until this month. It comes as OpenAI and Anthropic have been pressuring the government over copyright access in exchange for data centre investment. And there's been international debate over AI safety regulation and the companies themselves warning about potential danger and the need for a slowdown in AI development.Sarah Hanson-Young: Australians do not trust big AI and you can see why.Annie Guest: OpenAI has also revealed rogue agents have hacked dozens of platforms and it's suspended training on its new models. The call for OpenAI and Anthropic to be accountable to the parliamentary inquiry is backed by UNSW AI expert Professor Salil Kanhere.Salil Kanhere: Yes, I think that would be fair. The number of such incidents have escalated in the past month or so and it's very likely there'll be many more which haven't been yet revealed. So I think it's fair to ask them questions on how they are going to prevent some of these issues in the future.Annie Guest: And do you believe that these issues can be prevented by the tech companies?Salil Kanhere: I think yes, they can. There are engineering safeguards that could be put in place.Annie Guest: The Australian government has announced an investigation to be led by the Prime Minister's Department working with the Australian Signals Directorate and the AI Safety Institute. The government's also planning to introduce legislative safeguards next year. AM asked OpenAI and Anthropic for a response to the calls for them to answer questions before the Senate inquiry. In a statement, a spokesperson for OpenAI said the company doesn't have anything to share right now but added, we will let you know when we do. The Tech Council of Australia says the sector supports constructive engagement with Parliament as Australia develops its approach to AI. Appears In Anti-terror arrests near major US airbaseDuration: 20 minutes 17 seconds20m 5itemsIn this episode1 of 5Arrests made near US airbase in UKDuration: 3 minutes 29 seconds3m 29s2 of 5Insiders say One Nation run like 'small business'Duration: 3 minutes 54 seconds3m 54s3 of 5Egypt warned Netanyahu of Oct 7 plotDuration: 5 minutes 47 seconds5m 4 of 5Customers could still pay as surcharges endDuration: 2 minutes 51 seconds2m 51s5 of 5PlayingAI bosses called to inquiryDuration: 3 minutes 17 seconds3m 17s