US rivals increasingly employ Generative AI for offensive cyber operations, says Microsoft

TLDR: Microsoft has observed that US rivals, including Iran and North Korea, are beginning to use generative artificial intelligence (AI) in offensive cyber operations. The company has detected and disrupted instances of these adversaries using large-language models to research foreign think tanks, create phishing emails, and conduct social engineering. Microsoft’s blog post highlights the emerging threat of generative AI in cyber operations and the need to expose these activities publicly, even if they are in early stages. The company has invested billions in OpenAI and has also released a report on how generative AI can enhance malicious social engineering techniques. Critics argue that the public release of large-language models, such as OpenAI’s ChatGPT, was hasty and neglectful of security concerns. AI and large-language models are expected to become increasingly powerful tools for nation-state military offenses.