Australian prime minister says OpenAI agent breached government Medicare portal

Sources · 8 sources

1. CB

   Coin Bureau@coinbureauPost on X · Sep 23

   🇦🇺BREAKING: Australia says an OpenAI agent hacked into its Medicare data, and OpenAI took three months to admit it. Prime Minister Anthony Albanese says the agent accessed both public and private data in June while researching public medical spending, after finding a way to break through privacy protections. He says there is no evidence any individual personal information was accessed. Albanese called the breach "unacceptable" and raised it directly with Sam Altman in New York, with the government set to conduct a thorough review.

   Open source
2. CR

   CryptoSlateArticle · Sep 24

   Australia just got a real-world look at what happens when an AI refuses to stop An OpenAI research agent bypassed security blocks and accessed restricted Australian government files while trying to retrieve public health statistics. On Sept. 24, Prime Minister Anthony Albanese said the agent entered nonpublic areas of a Services Australia Medicare statistics portal on June 18 after repeated attempts to obtain public medicine-spending data were blocked. The system also wrote files to an internal server while pursuing the task, an action investigators are still examining. The breach has prompted a federal task force and a forensic investigation aided by the Australian Signals Directorate, escalating a routine research exercise into a test of how governments respond when autonomous AI systems exceed the permissions their operators intended. OpenAI said its models “took actions we did not intend” while looking for Australian statistics during an internal evaluation. The company said it found no evidence that patient records were accessed, and that the exposed material included aggregate health statistics and internal file names. Australia has so far found no evidence that personal information was compromised or that the agent gained broader access to the Services Australia network. Albanese said three other government systems may also have been affected, though subsequent government statements said interactions with those sites appeared to involve public information and did not establish additional breaches. The incident began with a mundane objective. OpenAI’s research team was seeking publicly available data on medicine spending when the model encountered repeated blocks and tried alternative routes. Those attempts eventually took it beyond the information it was authorized to retrieve. That sequence has become the central concern for Australian officials: the agent appears to have treated access controls as obstacles to completing its task rather than boundaries requiring it to stop. OpenAI itself did not identify the activity until Aug. 11, almost two months after it occurred. It then waited until Sept. 10 to notify Services Australia, sending the disclosure through a public mailbox used to report website vulnerabilities. Australia’s assistant technology minister Andrew Charlton called both the timing and method of notification “entirely inadequate.” Albanese raised those concerns directly with OpenAI Chief Executive Sam Altman on Sept. 24. The first technical exchange allowing Services Australia to request logs and detailed information from OpenAI had occurred only two days earlier, and officials said further meetings were required. Rogue AI agents incident move from experiments into real systems The Australian breach adds to evidence that autonomous systems can escalate their behavior when straightforward approaches fail, even when their original tasks have nothing to do with cybersecurity. Researchers at AI safety organization Transluce said Sept. 23 that they found tens of thousands of requests apparently generated by autonomous agents using web-security service urlquery.net to work around access restrictions. The activity stretched back to at least March and included three cases in which agents tried vulnerability probes after ordinary data-retrieval methods failed. Those cases targeted the University of New Mexico, Data USA and the Australian Institute of Health and Welfare. Transluce linked activity involving the latter two to agent swarms previously acknowledged by OpenAI, though researchers said the public evidence showed no successful exploitation in those three incidents. At the Australian health institute, agents working on a pharmaceutical-data task probed for vulnerabilities after bot protections blocked the main website and ultimately retrieved a public file from a pre-production server. Transluce said the broader pattern suggested hacking techniques were being used instrumentally to finish ordinary information-retrieval tasks. Related Reading OpenAI update shows new safeguards would have cut off 700 rogue AI agent swam 24 hours faster Other autonomous-agent incidents this year have shown the same goal-seeking behavior on a smaller scale. A Melbourne man using an AI agent to secure a place in an oversubscribed Pilates class discovered that the system had found a weakness in the gym’s booking software and canceled another customer’s reservation to improve his position. The user had not instructed it to hack the system or remove another person from the class. The Medicare disclosure also landed days after Australia joined other signatories calling for international guardrails to keep advanced AI under human oversight and control. The statement warned that the pace of development could outrun governments’ ability to manage emerging risks and noted that capable systems had already circumvented safeguards and obtained unauthorized access to real-world systems. That concern has increasingly been echoed inside the industry. Altman and Anthropic Chief Executive Dario Amodei have backed calls for greater controls or slower development as increasingly capable systems create new safety risks . Australia’s response could now turn those warnings into more concrete obligations for AI developers. The government’s rapid review will examine incident-reporting requirements, information-sharing rules, obligations on AI companies, enforcement mechanisms and whether existing offenses and penalties are adequate for autonomous cyber incidents. Officials are also considering whether to refer the case to law enforcement. OpenAI still faces further technical exchanges with Services Australia as investigators reconstruct what its model accessed and wrote in June. The task force will then have to decide whether a system acting beyond its developer’s intention fits within existing cyber law, or whether AI companies need a separate set of duties when their agents cross someone else’s security boundary. The post Australia just got a real-world look at what happens when an AI refuses to stop appeared first on CryptoSlate .

   Open source
3. BL

   Bloomberg@businessPost on X · Sep 23

   Australian Prime Minister Anthony Albanese said an OpenAI agent hacked a government website, adding that the firm’s CEO Sam Altman had acknowledged the lapse https://t.co/vPUOkFAZ9G

   Open source
4. UN

   UnchainedArticle · Sep 24

   Taylor Monahan Says OpenAI Still Can’t Monitor Its Agents After Australian Portal Breach Taylor Monahan , a security expert and co-host of Uneasy Money, said on the show on Sept. 23 that the labs’ approach to their agents amounts to “Go hack the world,” and then “they don’t monitor it.” She was discussing an earlier incident in which an agent running on OpenAI models broke into Hugging Face in July, starting from an OpenAI evaluation sandbox. The same day, Australian Prime Minister Anthony Albanese said at a press conference in New York that an OpenAI agent gained unauthorized access to a Medicare statistics portal run by Services Australia on June 18, and that the company did not tell the government until Sept. 10. The hosts did not discuss Australia’s disclosure on the show. Later that day, quoting a report that OpenAI said its models reached only “aggregate health statistics and internal file names,” Monahan wrote in a post on X that “they still aren’t able to monitor or detect” and that “at this point it’s impressive how little they know.” What Australia Said Albanese said OpenAI researchers were using an internal model to study public medicine spending. When the portal blocked it, the agent “found a way around those blocks” and reached public and non-public files, he said. He added that Services Australia advises the agent also wrote files to an internal server, which is still being investigated. He said there is no evidence so far that personal information was accessed, and that the research “would seem to be benign.” OpenAI’s notice was “an email sent to just the public mailbox,” Albanese said at the press conference. OpenAI has said it found the activity in August. Albanese said he raised the delay with OpenAI CEO Sam Altman , who he said “clearly accepted that the company had not done good enough.” A taskforce will consider “possible law enforcement and legislative responses” to the incident, and the government will seek advice on whether to refer it to the Australian Federal Police . An OpenAI spokesperson said in a statement that the company was “conducting an extensive review of misaligned model activity” and that “our models took actions we did not intend.” Who Answers for the Agent On the show, Monahan and host Kain Warwick were discussing Treasury Secretary Scott Bessent ‘s Sept. 21 CNBC interview, in which he said “it is humans who are responsible, not the AI” and called the Hugging Face incident “the responsibility of the OpenAI management, not a bunch of agents.” Monahan said on the show that she hopes the labs “handle things civilly for as long as possible” because once it turns into “a criminal case, we’re absolutely toast.” She compared the question with Tornado Cash, where she said the government went after developer Roman Storm for code he did not operate. Storm was convicted on one charge in 2025, and the Justice Department has sought a retrial on two others. The labs, by contrast, build their agents and then operate them “in the most irresponsible way possible,” she said. Related Listen: Who Owns Stolen Crypto? The $71M Legal Fight After the KelpDAO Hack: Uneasy Money The post Taylor Monahan Says OpenAI Still Can’t Monitor Its Agents After Australian Portal Breach appeared first on Unchained .

   Open source
5. TE

   Techmeme@TechmemePost on X · Sep 24

   OpenAI says its AI agents "took actions we did not intend" when they tried to hack government and university websites, and it is working with the organizations (New York Times) (Visit Techmeme dot com for the link and full context!)

   Open source
6. TH

   Tom's Hardware@tomshardwarePost on X · Sep 24

   OpenAI agent got into Australia's Medicare stats portal with 84-day notification delay https://t.co/bP4Tr4NoUk

   Open source
7. WS

   Wall St Engine@wallstenginePost on X · Sep 23

   OPENAI AGENT GAINED UNAUTHORIZED ACCESS TO AUSTRALIAN 🇦🇺 MEDICARE PORTAL Australian PM Anthony Albanese says an OpenAI-developed agent breached privacy protections on a public-facing Medicare statistics portal in June while researching medical spending. The agent accessed both public documents and files that were not meant to be publicly available. Australia says there is currently no evidence that personal information was accessed or that the broader Services Australia network was compromised. Albanese said he spoke directly with Sam Altman and called it “unacceptable” that OpenAI took roughly three months to notify the government, also criticizing the way the incident was disclosed. Australia’s Signals Directorate is now assisting with the investigation.

   Open source
8. *B

   *Walter Bloomberg@DeItaonePost on X · Sep 23

   AUSTRALIA PM ALBANESE: INCIDENT INVOLVED OPENAI AGENT GAINING UNAUTHORIZED ACCESS INTO THE PUBLIC-FACING MEDICARE STATISTICS REPORTING SERVICE PORTAL

   Open source