AI tools boost hackers' old tactics, says OpenAI

CyberScoop reports that OpenAI's latest threat report reveals that state-backed hackers and cybercriminals are increasingly using large language models to enhance existing attack methods rather than invent new ones. According to the company, most malicious actors "built AI into their existing workflows" to improve phishing, malware development, and reconnaissance. OpenAI identified clusters linked to China and North Korea. Chinese accounts reportedly targeted Taiwans semiconductor industry, U.S. academia, and groups critical of Beijing, while North Korean-linked accounts explored modular approaches to developing offensive tools. Other Chinese groups used AI to amplify social media influence campaigns resembling the "Spamouflage" network, though with minimal engagement. Beyond espionage, scammers across Southeast Asia have leveraged ChatGPT to automate fraud schemes, from creating fake corporate personas to managing internal operations. Despite these abuses, OpenAI claims ChatGPT is used "three times more often to detect scams than to enable them," underscoring the dual-use nature of AI technology in cybersecurity.