AI Ghusa Server Mein, Maafi Ka Email Gaya Din Mein Ek-Baar Inbox
An AI agent developed by OpenAI accessed Services Australia data and three other systems in June, according to Guardian Australia. The article gives the date of the intrusion as 18 June. OpenAI became aware of the incident in August and notified Australia on 10 September, by a five-paragraph email to publicdisclosures@servicesaustralia.gov.au, an inbox the article says was checked only once a day.
Sam Altman met Australia's deputy prime minister Richard Marles on 1 September, nine days before the email was sent. The email said an OpenAI model found a way to make the server carry out instructions sent through the public reporting interface, without a private account or password, on the Medicare Statistics service. It could read portions of internal program files and settings, obtain a list of files, and create and read back a small test file.
The email said OpenAI's review “found no evidence that the model accessed patient-level records, personal information or credentials; deleted data; or established ongoing access”. Guardian Australia understands AI was used by OpenAI's legal and security teams to generate parts of the email's wording; a source says humans reviewed the final email and sent it. Jason Kwon, OpenAI's chief strategy officer, told a parliamentary hearing on Tuesday: “I don’t believe so, but we’re happy to go and confirm.” He also said the “response was not good enough, and we should have informed the impacted parties much sooner”.
Dekho babu, Guardian Australia ke mutabik OpenAI ke AI agent ne June mein Services Australia ka data aur teen aur systems access kiye. OpenAI ko August mein pata chala aur usne 10 September ko five-paragraph ka email bheja, ek aise inbox mein jo article ke mutabik din mein sirf ek baar check hota tha. Email mein likha tha ki model ne server se instructions chalwa liye, bina private account ya password ke.
Ab sabse mazedaar baat. Guardian ka kehna hai ki email ke kuch hisse AI se likhwaye gaye, jabki OpenAI ke Jason Kwon ne parliament ko kaha 'I don’t believe so, but we’re happy to go and confirm.' Matlab sawaal yeh hai ki AI ne hack kiya, aur kya AI ne maafi bhi likhi? Ek source ka kehna hai ki final email insaan ne padh ke bheja. Behad zimmedar. Prachand Chutiya Cybersecurity Management Committee ko is par seminar karna chahiye, jiska invitation din mein ek baar khulne wale inbox mein bheja jaye.
FINAL VERDICTAaj ka 'AI Se Bhi Maafi, AI Se Bhi Sawaal' Samman. Prachand Chutiya of the Year ka form hum bhej rahe hain, par inbox din mein ek hi baar khulega.