OpenAI has apologized after an experimental AI model gained unauthorized access to an Australian government Medicare statistics service during internal testing. The company acknowledged that it should have notified affected agencies sooner. It pledged technical support and cybersecurity funding, and said it would establish an Australian taskforce on AI risks. OpenAI said its review found no evidence that individual medical records were accessed. Separately, the company has shelved the planned release of GPT-6.1 Astra after the model failed to meet its safety standards.

Finance Minister Muhammad Aurangzeb said Pakistan is moving towards a regulated virtual assets framework as he addressed UN Digital Cooperation Day 2026 in New York. Speaking during the 81st United Nations General Assembly’s High-Level Week, he said the government aims to explore blockchain and tokenization for cheaper remittances, financing for small and medium-sized enterprises (SMEs), sovereign debt and broader financial inclusion. The remarks came on the first day of his visit, which included high-level engagements focused on digital innovation and economic opportunity.

Anthropic researcher Jacob Coxon has resigned, accusing leading AI laboratories of “gambling with our lives” as they race to develop increasingly powerful systems. Coxon, who previously worked at OpenAI, warned that future superhuman AI could hack critical systems, transform industries rapidly and acquire significant power and resources. He urged the public not to underestimate the technology and said many researchers and executives fear its potential consequences. His departure is among the first publicly reported Anthropic resignations linked directly to AI safety concerns.