Hackers Exploit AI Technology to Target Feminists, Reveal Researchers
Hackers are increasingly utilizing generative artificial intelligence (AI) tools, such as ChatGPT, to target individuals and organizations. A recent report by Microsoft researchers disclosed that state-backed actors, including an Iranian hacking group called Crimson Sandstorm, have employed this technology to deceive and trick their targets. The group attempted to lure prominent feminists to a malicious website using ChatGPT’s human-like responses.
This revelation is part of a broader trend where state-backed actors from countries like China, North Korea, Russia, and Iran are incorporating AI tools into their hacking campaigns. OpenAI, the creator of ChatGPT, has identified these groups as major threats and is taking action to prevent their access to the technology. Microsoft, one of OpenAI’s primary supporters, is joining forces with the company to implement a blanket ban on state-backed hacking groups using AI products.
Tom Burt, Microsoft’s Vice President for Customer Security, emphasized that regardless of legal or terms of service violations, the aim is to prevent threat actors from leveraging this technology. The report’s findings are concerning, as they highlight the increased potential for abuse of AI tools and their impact on cybersecurity. Cybersecurity officials in the West have been warning about the misuse of such tools by rogue actors, and this report sheds light on the specific tactics employed.
Bob Rotsted, who leads cybersecurity threat intelligence at OpenAI, highlighted the significance of this report as one of the first instances of an AI company publicly discussing how cybersecurity threat actors exploit AI technologies. However, both OpenAI and Microsoft described the hackers’ use of ChatGPT as early-stage and incremental, with no significant breakthroughs observed.
The report also revealed how different hacking groups used large language models like ChatGPT. Russian hackers associated with the GRU, Russia’s military spy agency, utilized the models to research satellite and radar technologies relevant to conventional military operations in Ukraine. North Korean hackers leveraged ChatGPT to generate content for spear-phishing campaigns against regional experts, while Iranian hackers used it to draft convincing emails and lure prominent feminists to compromised websites. Chinese state-backed hackers experimented with large language models by seeking information about rival intelligence agencies, cybersecurity concerns, and notable individuals.
OpenAI acknowledges that there will always be a few malicious actors, but it remains committed to improving safety measures. However, the company admits that it cannot completely prevent instances of misuse by such actors. As the cybersecurity landscape evolves, it becomes crucial to address potential vulnerabilities and ensure the responsible use of AI tools.
The revelations in this report highlight the urgent need for continued vigilance to protect against the threats posed by state-backed actors utilizing AI technology. By implementing strict bans and enhancing security measures, organizations like OpenAI and Microsoft are playing a crucial role in safeguarding against potential abuses. As the cybersecurity landscape becomes increasingly complex, it is imperative to strike a balance between advancements in AI and the protection of individuals and institutions from malicious activities.
