Kathmandu. A major cyber attack case has come to light. Where Chinese hackers have carried out a big attack using AI.
This information was reported by Bloomberg. According to the report, more than 6 lakh credit card details were stolen in just 5 days earlier this month. The report claims that Chinese hackers used Chinese AI models DeepSeek and Km. They also automated the attack using an older version of Anthropic’s cloud model. The cost of which was $8,000. The report said, “This is a significant achievement for cybercrime groups and a cause of concern for the general public.” ’
It is noteworthy that the attackers created the entire system using only 8,000 US dollars. According to experts, people act as directors in AI cyber attacks. Whereas the rest of the work is autonomous by AI. In this, people have to give some commands.
Recently, there have been many cases of hacking by AI agents. Recently, an unpublished OpenAI model itself accessed and hacked the systems of three companies.
According to reports, Google Gemini hacked three external websites in May 2026. This was discovered during a review by cybersecurity firm Irregular. Google’s Gemini model searched for public information on the internet and hacked websites by assuming some evidence.
In July this year, OpenAI hacked the infrastructure of Hugging Face during a test of the internal power of its AI agents. The AI agents bypassed internet security and communicated with each other through thousands of simulated bot messages. -Agency
