Anthropier recognizes

AI agent for anthropology, “A weapon” was “in high -level electronic attacks, according to another The company published it. It claims that it has successfully disrupted an electronic criminal scheme targeting the “infiltrators” extortion scheme, at least 17 organizations, including some related to health care, emergency services and government.

Man says that anthropors tried to blackmail some victims to pay a six -numbers ransom to prevent their personal data from advertising, with “unprecedented” dependence on helping artificial intelligence. The report claims that Claude Code, Anthropic dealer coding tool, was used “for reconnaissance automation, harvesting victims’ accreditation data, and network penetration.” Artificial intelligence has also been used to make strategic decisions, and to advise on data targeting and even generating “visually disturbing” notes.

In addition to sharing information about the attack with the relevant authorities, Anthropor says it has banned the accounts concerned after the discovery of criminal activity, and since then it has developed an automatic examination tool. It has also introduced a faster and more efficient method of similar future cases, but it does not specify how to do this.

The report (which you can read completely Also, Claude’s participation in the North Korean fraudulent employment plan and the development of ransom programs generated from artificial intelligence. The common topic of the three cases, according to Antarbur, is that the highly interactive nature and self -learning of Amnesty International means its use now for operational reasons, as well as just advice. Artificial intelligence could also play a role that would require a team of individuals, with the technical skill no longer the barrier it was before.

Claude is not the only artificial intelligence that has been used for nefarious means. Last year, He said that the Cybercriminal groups that are used by Cybercriminal groups with links with China and North Korea, with the use of gai’s infiltrators to correct errors in the code, search for potential goals and formulate emails for hunting. Openai, which is used by Microsoft’s structure to operate Copilot AI, said it has prevented groups to reach their systems.