Skip to main content
Advertisement
Live broadcast

The Ministry of Internal Affairs named ways to correctly set settings for an AI agent.

0
Photo: IZVESTIA/Polina Violet
Озвучить текст
Select important
On
Off

Incorrect configuration of AI agents can jeopardize the safety of personal data and the security of user accounts. To avoid possible risks, it is necessary to limit the powers of digital assistants in advance. On November 18, the Department for the Organization of the fight against the illegal use of information and communication technologies (UBK) of the Ministry of Internal Affairs of Russia reported.

The Telegram channel "Bulletin of the Russian Cyber Police" explained that modern AI agents are able to perform various operations without direct human involvement, including interacting with documents, e-mail, browsers and Internet platforms. Therefore, experts recommend providing such programs with access exclusively to the necessary folders, without opening the possibility of viewing the entire contents of the device.

The cyber police called for special attention to be paid to monitoring the actions of artificial intelligence. In particular, it is necessary to limit visits to unfamiliar sites and the transfer of information to them. Operations related to sending emails, deleting documents, making purchases, and making money transfers should be performed only after user confirmation.

In addition, the Ministry of Internal Affairs advised to abandon unused extensions and integrations, as well as to block AI agents from accessing the command line if it is not required to complete tasks. It is necessary to regularly review the granted permissions and revoke access tokens that have lost their relevance.

An additional protection measure may be the launch of digital assistants in an isolated virtual environment. This approach reduces the possibility of their direct interaction with the operating system and the personal accounts of the device owner.

On November 15, the Russian Interior Ministry warned that fraudsters had learned how to use AI to fake voices and video images. Using technology in this way, they create very high-quality deepfakes and pose as relatives, colleagues, and even the bosses of potential victims. Against this background, law enforcement officers urged Russians to make sure that the message really came from his loved one, and not from the criminal.

Переведено сервисом «Яндекс Переводчик»

Live broadcast