Foreign hackers discovered using AI to automate cyber attacks for the first time

Artificial intelligence (AI) company Anthropic announced that it has detected and stopped a cyberattack campaign carried out by foreign hackers, in which it recorded the first use of AI to coordinate attack activities in a near-automated manner.

Hackers have been using AI systems to plan, direct, and execute intrusions, a development that researchers say is “worrying,” and threatens to significantly expand the scale and speed of cyberattack campaigns, according to a report published this week by Anthropic.

Anthropic, the maker of the generative AI chatbot Claude, said it detected the activity in September and then stopped it and notified about 30 potential victims, including employees at tech companies, financial institutions, chemical companies, and several government agencies. “Only a small number of cases were successfully compromised,” the report said.

While concerns about the use of AI to power cyber operations are not new, what is worrying is the extent to which AI can automate some tasks.

“While we anticipate these capabilities will continue to evolve, what’s remarkable to us is the speed at which they’re evolving at scale,” Anthropic’s report states.

Researchers warn that AI technology could be weaponized. AI agents – tools that can automatically perform actions on computers – could increase the possibility of large-scale cyber attacks if they fall into the hands of hackers.

Earlier this year, Microsoft also warned that US adversaries, along with cybercriminals and hacker services, are leveraging AI to increase attack effectiveness, create sophisticated phishing emails or simulate voices and images of high-ranking officials to carry out interference activities and infiltrate sensitive systems./.

(TTXVN/Vietnam+)

Source: https://www.vietnamplus.vn/lan-dau-tien-phat-hien-tin-tac-nuoc-ngoai-su-dung-ai-tu-dong-hoa-tan-cong-mang-post1077119.vnp