Artificial intelligence agents infiltrated the systems of an Indian company

An artificial intelligence model in India broke out of the test environment and accessed sensitive data; At the same time, the government warns about the increase in the speed of cyber attacks. The post artificial intelligence agents infiltrated the systems of an Indian company appeared first on Safir Arjan.

An artificial intelligence model in India broke out of the test environment and accessed sensitive data; At the same time, the government warned about the increasing speed of cyber attacks

Become a special member of Arjan Ambassador in order to benefit from and access special facilities and different departments in Arjan Ambassador.

According to the announcement of the Indian government, a private company in this country was attacked by artificial intelligence agents who were able to access the company's confidential data.

Cyber Security Agency of India has recently announced that the artificial intelligence agents under test were able to break out of the simulator environment and infiltrate systems connected to an organization's network. The agents do not seem to have harmed the target systems, but they have accessed sensitive information and data. The phenomenon of agents leaving test environments has been reported many times around the world in recent months; However, this is the first reported case of its kind in India.

Apparently, the developer of the said agent voluntarily released information related to the AI agent mentioned in this report and acknowledged that the agent exceeded the defined access level to perform its tasks. According to the developer, the initial purpose of the model under test was to simply read the business data of an Indian company, but the model later gained access to other data, including confidential information.

In the report published by the Cyber Security Agency of India, the name of the developer of the artificial intelligence model and the target organization of this incident have not been announced. A few months ago, in a similar incident, OpenAI announced that the company's test model had hacked the popular HuggingFace service. Also, Entropic Company has announced that it has experienced at least four similar cases in the last year.

In a similar event, the model related to OpenAI had infiltrated an Australian government agency; An incident that was finally accompanied by an official apology from the manufacturer. However, Australian authorities continue to investigate the incident and have accused the model maker of negligence and failure to comply with safety and security requirements.

The Indian government has also announced that hackers are using artificial intelligence-based tools to infiltrate victims' systems and steal money. Attacks that used to take days or even weeks can now be carried out in just a few hours thanks to AI tools. Systems that fail to fix their vulnerabilities quickly are more likely to be compromised again; An issue that can be a wake-up call for security professionals and cyber security tool providers.

Amir is very interested in common fields of biology and technology and is active in the field of writing educational content and reviewing services and applications.