In the United Kingdom, parliamentarians said that the country’s Prime Minister Rishi Sunak’s plan to make the UK a leader in the regulation of artificial intelligence is under threat.

According to ministers from the House of Commons Technology Committee, the European Union may overtake the United Kingdom in the race for the role of the main regulator of the development and dissemination of AI. According to them, in order to avoid this scenario, the UK should take measures consisting of the entry into force in November of a new law creating a legal basis for control over the technology sector.

An international artificial intelligence summit will be held in Great Britain in early November. The Government of the United Kingdom informed reporters of its readiness to take additional measures to ensure the country’s role as a leader in regulating advanced technology. But at the same time, this statement does not contain comments on the possibility and likelihood of the adoption of a new law.

The UK will allocate an initial investment of 100 million pounds to a task force to promote the safe development of machine intelligence models. The government of the country stated that this amount of financial injections in ensuring the security of AI exceeds government spending for similar purposes in any other country in the world.

The Technology Committee of the House of Commons, in a report published on Thursday, August 31, said that if the new legislation is not presented in the King’s speech on November 7, the earliest date for the introduction of legal innovations may be 2025.

The aforementioned report also indicates that in the absence of measures to adopt a new law over the next two years, the UK will not become a leader in the area of artificial intelligence regulation, since the EU AI Act on Artificial Intelligence, against the background of the lack of proposals from the shores on the other side of the English Channel, has every chance of obtaining the status of a kind of common standard, which will be extremely difficult to replace. The Technology Committee of the House of Commons justifies its position by the fact that there is already a similar example in the history of legal initiatives, which consists of the fact that the legislative framework of the United Kingdom in the field of data protection inherits the EU concept in many ways.

The government’s white paper on the regulation of artificial intelligence mentions that at some point a new law may become a necessary measure, without which Great Britain will not be able to become the leader of AI control. But at the same time, Rishi Sunak stated that within the framework of these efforts, much can be done without legislation.

The Prime Minister’s plan to ensure the main role of the United Kingdom in the regulation of machine intelligence provides that the November summit, which will be the world’s first global meeting on the security of advanced technology, is one of the main stages on the way to the goal.

The Technology Committee of the House of Commons stated that as many countries as possible, including China, should take part in this summit.

The committee report also pays special attention to issues that need to be addressed by the Government of the United Kingdom. One of the tasks is to overcome bias. For example, AI in the process of securing employment can associate female names with professions that, within the conservative paradigm, are designated as the most suitable for representatives of this sex. Because of this problem, job offers for women can be limited.

The Committee draws attention to the issue of confidentiality. Machine intelligence tools can be used to identify people in contradictory ways. For example, the police apply real-time facial recognition systems that scan the external data of citizens and compare them with lists of suspects.

The report also indicates that machine intelligence systems can replace some jobs. In this context, the Government should analyze the economic consequences of the spread of a new generation of technology.

Another problem is the training of artificial intelligence systems based on materials that are protected by copyright. Generative AI can independently create original works in the style of famous painters, inherit the creative approach of musicians without plagiarizing, and reproduce the behavior of actors in a special way, but not copy their habits. To obtain these opportunities, it is necessary to train artificial intelligence systems based on a variety of copyrighted materials. Many actors, artists, writers, and musicians say that training should take place only with their permission and subject to compensation. The report contains information that measures are already being taken to conclude a voluntary agreement that will allow AI companies to gain access to materials protected by legislation, and at the same time support the authors.

Copyright exemption for firms working in the area of artificial intelligence was canceled by the British government in February.

Another problem is that the ability of AI to imitate people can become a tool for spreading disinformation. Members of parliament say that this function can be used for fraud, for example, to deceive bank security systems with voice recognition.

The report was released after a statement by the National Cybersecurity Center that large language models controlling chatbots cannot be protected from certain attacks aimed at forcing AI to perform malicious actions. The Center noted that currently there are no measures by which this risk can be eliminated.

In general, the deputies support the government’s concept of ensuring the security of machine intelligence, which provides not for the creation of a separate supervisory authority, but for granting new powers to existing regulators.

Some of the speakers on the committee, including Hugh Milward from Microsoft UK, called the British approach preferable and described the EU regulatory framework as an example of how not to do it. But Mr. Milward also told reporters about the need to be cautious about any initiative of the United Kingdom, noting the danger of attempts to make one law a universal legal instrument regulating an excessive number of aspects of the AI industry.

As we have reported earlier, SEC Chair Says About Developing Rules for AI Use.