In the United States, the government-backed nonprofit organization Mitre on Monday, March 25. She announced the start of the laboratory’s work to test artificial intelligence systems used by federal agencies.

As part of the mentioned activeness, machine intelligence configurations will be checked for flaws to eliminate them later. Attention will also be paid to the security of artificial intelligence systems. Moreover, these digital products will be tested regarding other potential risks.

Miles Thompson, a robotics engineer who will head the laboratory, said that artificial intelligence systems will be tested for various parameters, including, among others, the degree of probability of data leakage and explainability, which is the ability to understand why AI makes certain decisions.

The laboratory is based in McLean, Virginia. Mitre specializes in overseeing research in the areas of national security, aviation, cybersecurity, and healthcare. The physical presence of 50 people is possible in the laboratory. It also provides for the participation of 4,000 people through remote connections.

Experts note that artificial intelligence systems, which are often likened to black boxes, are being implemented without a full understanding of the many ways to deceive these AI configurations and the causes of failures in the process of their functioning.

As part of the rapid adoption of artificial intelligence by the United States government, Miles Thompson and his co-workers often receive requests to conduct a one-off assessment of advanced technology. Against the background of the emerging sustainable need for Mitre, it was decided to create a laboratory to handle the process of testing AI systems.

The new facility was called AI Assurance and Discovery Lab. The laboratory will learn the risks of artificial intelligence systems by hacking them and testing for bias.

Democratic Senator Mark Warner of Virginia, chairman of the Select Committee on Intelligence, was among the senior lawmakers who visited the new facility on its opening day. According to him, the laboratory’s activities are an attempt to maximize the benefits of artificial intelligence while reducing some of the risks associated with advanced technology. Mark Warner stated the need for a comprehensive approach to studying and unleashing the potential of AI. The senator said that he was looking forward to the discoveries and progress that the laboratory would be able to make.

It is worth noting that the active development and scaling of the use of artificial intelligence have actualized the problem of cybersecurity. In this case, the risks are not only related to the fact that advanced technology may malfunction or become a tool for implementing various manipulations. Scammers and hackers also have access to AI. Criminals use artificial intelligence as part of their activities, as a result of which the methodology of their illegal work becomes more sophisticated, and, for this reason, the possibility of its detection becomes more complicated. In the context of new technological challenges, user awareness is extremely important. For example, by entering an Internet search query such as How to know if my camera is hacked, anyone can find out about signs of unauthorized access to a personal device.

Also, in the context of the spread of artificial intelligence, the problem of deep fake has intensified. The degree of realism of AI-generated content is very high, which is why detecting deception becomes a difficult task even for those who are technologically literate.