The expert spoke about the threats of using AI and ways to protect against them.

Currently, 70% of Russian companies have implemented AI solutions. However, the rapid development of the neural network carries not only advantages, but also serious risks. Maxim Smirnov, Deputy CEO of IVA Technologies, told Izvestia on June 29 about the threats AI poses and how to protect yourself from them.

"When neural networks were just beginning to develop, society was gripped by fears and warnings. Many people believed that artificial intelligence would become a threat to jobs, displacing people from various professions. This fear was justified: automation and new technologies are really capable of changing the labor market," the expert said.

He clarified that personal data leaks and deepfake fraud are among the main risks. In Russia, 19 major data breaches occurred in the first two months of this year, affecting 24 million people. To protect your data, it is recommended to avoid posting it in open sources and use encryption. For the corporate segment, it is preferable to install AI in a closed company circuit.

Smirnov drew attention to the fact that deepfake fraud is also becoming a serious problem: experts predict that by 2025, every second Russian will face such attacks. Currently, every fifth Russian company has been affected by the use of fake audio and video recordings. To avoid deception, it is important to double-check information sources and minimize the publication of personal data in the public domain.

In addition, modern neural network models do not have autonomous consciousness and act on the basis of training data, which can lead to bias. For example, Amazon's recruitment algorithm discriminated against women because of the dominance of men in the training data.

According to the expert, several steps are needed to reduce such problems. First, companies should be transparent and publish information about how AI works. Secondly, the data must be carefully checked for bias. The third important aspect is government regulation of AI.

"To increase confidence in AI technologies, it is critically important to create and use domestic solutions. Dependence on foreign platforms in key industries can turn into a strategic risk," the source concluded.

Transition to faces: authorities will start blocking fraudulent content created by AI

Will the new measures help solve the problem with deepfakes?

On June 25, it was reported that AI is rapidly entering the professional lives of Russians: almost half of the respondents (48%) are already using AI tools at work. The majority of citizens (81%) see new career opportunities in artificial intelligence. Most often, technology is used to prepare texts. AI is most in demand in design, marketing, and IT.