English
Saanchi Gupta
Saanchi Gupta is a dedicated trainee journalist at India.com English Desk, where she covers National, International, Viral, and Entertainment news. She has worked on major developments, including the ... Read More
Anthropic’s AI chatbot: As it’s said that technology is both a boon and a bane, some surprising news has come in, which states that a hacker allegedly used the AI chatbot of Anthropic, Claude, to breach the systems of the Mexican government. According to reports, the researchers mentioned that the activity started in December and kept happening for almost a month. The heavy data of 150 gigabytes was reportedly stolen, as reported by Bloomberg. This included documents associated with 195 million taxpayer records, voter records, and credentials of the government employee, along with the civil registry files.
According to the researchers, the hacker allegedly wrote Spanish-language prompts for instructing the AI chatbot Claude to act as a top-class hacker. The tool was reportedly utilised to identify the weak points and vulnerabilities in the networks of the Mexican government and extract data.
According to Gambit Security, the federal tax authority of Mexico and the national electoral institute were the most affected ones. In addition, the state governments like Jalisco, Michoacán and Tamaulipas, and many others were reportedly affected.
However, the tax authority of Mexico has denied such allegations, as reported and has said that no such breach has been found. Alongside, the national electoral institute has also denied the unauthorised access.
The Chief Strategy Officer at Gambit Security, Curtis Simpson, added, “In total, it produced thousands of detailed reports that included ready-to-execute plans, telling the human operator exactly which internal targets to attack next and what credentials to use.”
At the attack, Anthropic said, “We have blocked the problematic account and will incorporate this abuse case into our model training. Our latest models have enhanced safeguards against such misuse.”
When the hacker came across the alleged obstacles while using Anthropic’s AI chatbot Claude, the researchers, as stated by reports, mentioned that the attacker tried to use ChatGPT to fetch additional guidance. In addition, this included advice on the detection of risk, as per reports.
The cybersecurity experts warned that artificial intelligence is being used in ways unsaid. It includes the alleged breach of data by the hackers.
When AI is rapidly growing, it’s essential to remain updated with the potential vulnerabilities of such systems.
