5OpenAI Blog·
OpenAI bans Korean-language accounts using AI for malware development and phishing
OpenAI identified and banned accounts using its models to support malware development, debugging, phishing, and credential-theft workflows, with the activity conducted in Korean. The disclosure is part of OpenAI's ongoing series of threat-actor disruption reports. This is a concrete example of AI models being weaponized for offensive cyber operations and the countermeasures being applied.
Related guides (3)
Related events (8)
OpenAI bans accounts linked to Russian-speaking criminal groups building AI-assisted malware
OpenAI identified and banned accounts likely associated with Russian-speaking criminal groups that were using AI to develop malware loaders, evasion layers, credential-theft scripts, and command-and-control infrastructure. The action is part of OpenAI's ongoing effort to disrupt malicious uses of its platform. This case documents a concrete instance of threat actors operationalizing frontier AI for offensive cyber tooling.
OpenAI bans DPRK-affiliated accounts using AI for cyber intrusion research
OpenAI terminated accounts potentially linked to North Korea-affiliated threat actors who were using AI to research intrusion tooling, phishing campaigns, malware development, and cryptocurrency targeting. The action is part of OpenAI's ongoing effort to disrupt malicious uses of its platform by state-affiliated actors. This represents a concrete case of AI being weaponized for offensive cyber operations by a nation-state actor.
OpenAI bans Russian-speaking accounts using AI for malware development (Operation ScopeCreep)
OpenAI terminated Russian-language accounts that were using its models to develop malware, refine loaders, and troubleshoot offensive cyber tooling. The operation, dubbed 'ScopeCreep,' represents a disclosed enforcement action against AI-assisted malicious cyber activity. This is a concrete case of AI misuse for cyberweapon development being detected and disrupted by a frontier lab.
OpenAI bans accounts linked to PRC-affiliated threat groups using AI for phishing and scripting
OpenAI terminated accounts whose activity overlapped with publicly reported threat groups displaying hallmarks consistent with PRC intelligence requirements. The banned accounts were using AI to support phishing campaigns and scripting workflows. This is part of OpenAI's ongoing effort to detect and disrupt malicious use of its platform by state-affiliated actors.
OpenAI bans accounts linked to AI-assisted online fraud and scam networks
OpenAI identified and banned accounts associated with online fraud networks that were using its models to generate scam scripts, impersonate individuals, translate content, and engage with victims. The action is part of OpenAI's ongoing effort to disrupt malicious uses of AI. This represents a concrete enforcement case documenting how threat actors are operationalizing LLMs for fraud at scale.
OpenAI bans accounts linked to deceptive AI-assisted fake IT worker employment scheme
OpenAI banned accounts associated with suspected deceptive employment campaigns that used AI tools to generate materials for fraudulent remote-job applications. The operation appears connected to the broader pattern of North Korean-linked IT worker schemes that have targeted Western companies. OpenAI's disclosure is a primary-source safety enforcement action documenting a novel misuse vector.
OpenAI bans accounts linked to North Korea-style deceptive IT-worker employment scheme
OpenAI banned accounts suspected of facilitating a deceptive employment scheme bearing characteristics of publicly reported North Korea-linked IT-worker activity. The operation appears to involve using AI tools to support fraudulent job applications or remote work impersonation. This is a concrete enforcement action by a major AI lab against state-linked misuse of AI systems.
OpenAI bans Iran-linked STORM-0817 accounts using AI for malware development and scraping
OpenAI identified and banned accounts linked to the Iran-affiliated threat actor STORM-0817, which was using OpenAI's models to debug Android malware, scrape social media platforms, and translate offensive tooling. The disclosure is part of OpenAI's ongoing threat intelligence reporting on state-linked misuse of AI systems. This case illustrates concrete adversarial use of frontier AI for cyberoffense and surveillance tooling.