United Russia, the country's ruling, Kremlin-backed party, has proposed restrictions on the use of artificial intelligence (AI) in healthcare, education and other areas that it considers sensitive, claiming that the tech could be used in a harmful way.

In a recently released document called The Digital Manifesto, United Russia suggested that the government should ban decision making by AI in several areas, including healthcare, justice, education, as well as property and labour rights.

Apparently, United Russia's top brass are unhappy about the fact that in some crucial areas the decision-making authority has already shifted from humans to AI.

According to an anonymous source in the party quoted by RBC business daily, United Russia senior members are specifically concerned that AI is used in making decisions on issuing bank loans and for the preparation of some court documents.

"True, at this point, AI can't yet make legal decisions, but it may not be too far away, as it is similar to loan decision making, in which AI originally assisted before starting to make its own decisions," the source was quoted as saying, adding that Herman Gref, head of Russia's largest, state-run lender Sberbank, had allegedly admitted to his inability to explain decisions made by the new tech.

The source also pointed to recruitment decisions made by AI, saying that job applicants may not event be aware that their qualifications are evaluated by a robot.

Meanwhile, the involvement of AI in making loan decisions by Russian banks has been on the rise. Last year, Alexander Vedyakhin, Sberbank's first chairman, said that all loan decisions made by the lender involve AI, while 95% of decisions are fully automated, with no humans involved.

When it comes to legal matters, AI is used by Russian court for preparation and verification of court documents in a bid to save judges' time.

In late 2020, several companies affiliated with Sberbank launched a diagnostic service based on AI. Patients are expected to describe their symptoms in detail, and an AI-based system then establishes a diagnosis with a probability ranging between 75% and 91%.

Most recently, AI has been used for detecting coronavirus (COVID-19) symptoms as well as signs of cancerous tumours.

Andrei Turchak, secretary of United Russia's general council, defended the idea of curbing the use of AI in some vital areas, saying that the issues of digitalisation and use of AI are at the top of the party's agenda.

"At some point, we facilitated legislative decisions that now make it possible to introduce experimental legal regimes in some areas in order to enable the use of artificial intelligence technologies," he said in a statement sent to Russian media.

"We allowed organisations developing tech solutions of that kind to use anonymised data sets, including those from state information systems," he went on to say. "But the use of modern tech should in no way be harmful to people. There should be clear-cut mechanisms for protection of people when digital technologies and artificial intelligence are used."

According to Turchak, there should be guarantees that people's personal data used by AI systems should be carefully protected and should not be available to unauthorised parties.