State-sponsored hackers exploit AI tools for fraud and disinformation
OpenAI shut down ten networks of state-sponsored hackers from China, Russia, North Korea, and Iran exploiting ChatGPT for fraud, disinformation campaigns, and cyberattacks. The hackers used AI to create fake job recruitment scams across multiple languages and false social media campaigns impersonating American veterans to spread divisive content and destabilize political discourse.
What happened
OpenAI detected and shut down ten large networks exploiting its AI tools over a three-month period. State-sponsored hackers from China, Russia, North Korea, and Iran reportedly used AI as a weapon in digital warfare to conduct sophisticated campaigns threatening global security and internet trust.
Chinese hacker groups launched campaigns spreading false news and divisive messages on social media platforms including TikTok and X through operations named 'Sneer Review' and 'Uncle Spam', creating fake profiles often impersonating American veterans to destabilize social debate and political dialogue. North Korean hackers used AI to create fake biographies and job applications worldwide, likely to gather confidential information and bypass security barriers to access sensitive IT systems. Russian hacker groups developed malware with AI assistance, planned cyberattacks, and conducted operations including 'ScopeCreep' aimed at data theft.
The 'Wrong Number' group based in Cambodia used AI to translate and personalize messages promising easy earnings, offering high-paying jobs for simple activities like social media engagement before soliciting money for alleged rewards. This fraud operated in multiple languages and directed victims to WhatsApp and Telegram applications.
Reported impact
The campaigns manipulated information and created deep societal divisions, causing fear and distrust. AI-enabled cyberespionage and data theft posed serious threats to global cybersecurity.
Index analysis
The incidents demonstrate how state actors and criminal groups have adapted AI tools to scale traditional cyberattacks and fraud schemes across borders and languages, exploiting the technology's capabilities for multilingual communication and personalization.
Governance controls
OpenAI deployed its own AI tools to quickly detect and stop malicious activities and shut down the exploiting networks.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.