Microsoft identifies state-backed hackers using OpenAI tools

Hackers using OpenAI tools: State-backed hackers from Russia, China, and Iran have reportedly been using tools developed by OpenAI, a company supported by Microsoft, to refine their hacking techniques and deceive their targets, as disclosed in a report published on Wednesday.

According to Microsoft, it has identified hacking groups associated with Russian military intelligence, Iran's Revolutionary Guard, and the governments of China and North Korea employing large language models, commonly referred to as artificial intelligence (AI), to generate responses that mimic human language. Microsoft's findings prompted the company to implement a comprehensive prohibition on state-backed hacking groups accessing its AI products.

Tom Burt, Microsoft's Vice President for Customer Security, underlined the company's stance on restricting access to its technology for identified threat actors, irrespective of legal violations or breaches of terms of service. In an interview with Reuters preceding the report's publication, Burt highlighted Microsoft's commitment to mitigating potential risks associated with the misuse of AI technology by such entities.

Diplomatic officials from Russia, North Korea, and Iran did not immediately respond to requests for comment regarding the allegations. China's US embassy spokesperson Liu Pengyu countered the accusations, asserting China's opposition to baseless accusations and advocating for the responsible deployment of AI technology to benefit humanity.

The revelation of state-backed hackers leveraging AI tools to augment their espionage capabilities underscores concerns surrounding the widespread adoption of AI and its susceptibility to exploitation. Cybersecurity experts in the West have been cautioning about the misuse of such tools by rogue actors, although specific instances have been scarce until now.

Bob Rotsted, who oversees cybersecurity threat intelligence at OpenAI, noted that this is among the first instances of an AI company publicly addressing how cybersecurity threat actors exploit AI technologies.

Both OpenAI and Microsoft characterised the hackers' use of their AI tools as preliminary and incremental, with no reported significant advancements in their capabilities. According to Microsoft, the identified hacking groups employed large language models for various purposes, including reconnaissance on military technologies, crafting spear-phishing content, and composing convincing fraudulent emails.

"This technology is both novel and immensely powerful," Burt remarked, underscoring the need for vigilant oversight in its utilisation.

(With Reuters inputs)