
Australian Prime Minister Anthony Albanese warned at the UN General Assembly that AI is advancing too quickly without safeguards, citing an OpenAI agent's breach of an Australian government Medicare statistics portal in June, which was not reported to authorities until September.
AI-generated summary
Albanese addressed the UN General Assembly to call for international cooperation on AI regulation following a cybersecurity incident involving an OpenAI agent accessing Australian government data.
Australian Prime Minister Anthony Albanese warned that artificial intelligence is evolving at a “furious pace” and has urged other countries to shape its development, after an OpenAI agent breached an Australian government website this year.
Addressing the United Nations General Assembly in New York, Albanese called for international cooperation on emerging technology, adding that AI company leaders themselves have warned about advancing frontier AI too quickly without safeguards.
“Recently, an artificial intelligence agent infiltrated an Australian government website. This is unacceptable,” he told world leaders on Thursday.
The warning came after Albanese revealed that an OpenAI agent had breached a Medicare statistics portal in June, accessing public and non-public files.
Albanese said OpenAI did not notify the Australian government until Sept. 10, nearly three months after the incident.
OpenAI told Cointelegraph it became aware of the incident in August and investigated what information had been accessed before notifying Services Australia.
AI outlook — possibilities, not facts
Australia will push for stricter international AI accountability measures at future UN forums.
Likely · Within months

Researchers at Alloc Init proposed Shielded Bitcoin, a system to enable private Bitcoin transfers using encrypted notes and zero-knowledge proofs without requiring a protocol soft fork. The design leverages Bitcoin as a publication layer while relying on external indexers for validation, drawing both support and criticism over privacy limitations and quantum resistance.

Researchers from [[alloc] init] have proposed Shielded Bitcoin, a metaprotocol enabling private transfers on Bitcoin's network without requiring a soft fork or consensus changes, using techniques like zero-knowledge proofs and encrypted notes while leveraging Bitcoin's mainnet for data publication and ordering.

Block has joined the x402 Foundation and integrated Bitcoin Lightning into the x402 payment protocol to enable low-cost, high-volume transactions for AI agent commerce, aligning with Google, Microsoft, AWS, Coinbase, and Solana Foundation in supporting open standards for autonomous machine-to-machine payments.

A vulnerability in older Lightning Terminal software could incorrectly record Bitcoin Lightning invoices as paid after payments were canceled and returned to senders, potentially leading merchants to release goods or credit without receiving funds, Lightning Labs disclosed on Sept. 21, 2026; fixes were shipped prior to disclosure.

More than $183 million in cryptocurrency assets were moved from wallets linked to exchange Bitget to a newly created address in what analysts describe as a potential hack, with the first transaction involving $19.67 million in USDT0 swapped for ETH via UniswapX and 1inch Fusion; Bitget has not confirmed the exploit, and outflows ceased after approximately six minutes, consistent with an exchange freezing withdrawals during investigation.

A proposed class action lawsuit alleges OpenAI routed real user ChatGPT chats to outside contractors via Project Lily without proper disclosure, violating privacy laws by allowing human reviewers to access personal details despite automated filters, with plaintiffs seeking damages and injunctive relief including opt-in consent requirements.