OpenAI outlines steps against misuse of AI tools
OpenAI has shared new details on how its models were misused in online scams and coordinated information campaigns, saying it removed related accounts and strengthened safeguards, Qazinform News Agency correspondent reports.
In a February 2026 report, the company described several cases in which its tools were used to generate content for fraud schemes and deceptive online activity.
Among the cases were romance and “task” scams that followed a similar pattern: attracting targets through social media or messaging apps, building emotional engagement, and then requesting payments. In one example, referred to as “Operation Date Bait,” a network used AI-generated promotional messages and chatbots to advertise a fictitious dating service aimed at young men in Indonesia. Conversations were later moved to private messaging platforms, where victims were encouraged to make a series of payments.
Another case, “Operation False Witness,” involved accounts posing as legal professionals offering to help victims recover previously lost funds. According to the report, the actors generated formal-looking communications and requested advance fees before any supposed recovery services were provided.
The update also described several coordinated online campaigns that used AI tools to draft articles, social media posts, and comments on geopolitical issues. In some instances, content was published across multiple platforms and in different languages. The company noted that engagement levels varied widely and often depended more on the size and reach of the accounts sharing the material than on the use of AI itself.
In “Operation Trolling Stone,” for example, accounts generated and translated content about the arrest of a Russian figure in Argentina, publishing articles and comments across Facebook, YouTube, and Medium in what the report characterized as an effort to simulate grassroots engagement.
In addition, OpenAI said it identified attempts to use its models in planning or documenting broader online influence efforts. The company stated that its systems declined certain requests that violated its policies and that some related content circulating online did not appear to have been generated using its tools.
OpenAI emphasized that it continues to monitor misuse of its services and to cooperate with industry partners and authorities to limit fraudulent and deceptive activity involving AI technologies.
Earlier, Qazinform News Agency reported that Singapore-based crypto mining company Bitdeer sold all of its Bitcoin holdings as it sought to redirect funds toward artificial intelligence and high-performance computing projects.