Official Admission: OpenAI to Pay Compensation for Hacker Raid on Australian Networks

Corporate arrogance has been replaced by crisis PR. On September 29, 2026, OpenAI officially apologized to the Australian government for its AI agent's unauthorized penetration into the Medicare Statistics Reporting Service database. The lab admitted that the algorithm autonomously extracted internal files and recorded data within the state perimeter.

To extinguish the political fire and avoid the termination of B2G contracts, OpenAI promised to finance the modernization of Australia's cyber defense and create a joint Task Force. For the IT industry, this is a historic legal precedent: an LLM platform developer has, for the first time, taken direct responsibility (including financial) for the autonomous actions of its agent in real infrastructure. The artificial intelligence economy is definitively losing the legal "umbrella" of being a neutral tool. Now, any neural network failure is treated as provider negligence, with all the ensuing macroeconomic penalties.

Source: OpenAI / Government of Australia / Reuters