AI News Digest for the first half of 2024

In Russia

• AI Consultant on State Services Portal Gosuslugi

This year, Russia's Ministry of Digital Development plans to introduce GPT-based artificial intelligence technology to assist users of the federal electronic service Gosuslugi. The chatbot will not only suggest services based on the user's situation but will also automatically fill out necessary applications using the data stored in the system during the conversation.

According to media reports, the new technology is currently being tested in a closed environment, with the first results expected to be released in September 2024.

The AI consultant will utilize a domestic technology similar to YandexGPT and GigaChat, instead of the ChatGPT.

• Requirement of insurance for damages caused by AI and establish liability for such damages

In July of this year, amendments were made to the federal law of July 31, 2020, No. 258-FZ "On Experimental Legal Regimes in the Field of Digital Innovations in the Russian Federation" (hereinafter referred to as Law No. 258-FZ).

The purpose of these amendments is to reduce the risks associated with the use of AI technologies, specifically:

1. expanding the circle of persons responsible for causing harm to the life, health or property of a person or the property of a legal entity during the implementation of an experimental legal regime, including as a result of the use of solutions created using AI;
2. expanding the range of persons responsible for causing harm to life, health, or property of individuals or the property of legal entities when implementing an experimental legal regime, including as a result of using solutions created with the application of AI;
3. ensuring access to information about persons responsible for causing harm as a result of the use of technologies created with the application of AI and maintaining records of such technologies;
4. regulating the procedure for using technologies created with the application of AI, as well as ensuring conditions for civil liability insurance.

The amendments were made by Federal Law No. 169-FZ "On Amendments to the Federal Law 'On Experimental Legal Regimes in the Field of Digital Innovations in the Russian Federation,'" which was published on July 8, 2024, and will come into force on January 5 of the following year.

Among other things, according to the amendments to Law No. 258-FZ, legal entities and individual entrepreneurs participating in experimental legal regimes in the field of digital innovations will be required to insure against liability for harm caused to the life, health, or property of others during the implementation of the experimental legal regime, including as a result of using solutions developed with AI technologies.

It should be emphasized that the law contains important provisions that damage caused to the life, health or property of a person or the property of a legal entity during the implementation of an experimental legal regime is subject to compensation, even if it was caused by the lawful actions of a participant in the regime.

Under the new rules, the regulator will commission a review of all cases of harm that will be caused when using solutions developed using AI technologies. At the same time, the commission, in addition to representatives of the regulatory body, may include other interested parties, including representatives of organizations of the business and expert communities.

In addition to the actual volume, nature and circumstances of the harm, including technological failures and errors that were made during the development and implementation of AI technology, the commission must give an opinion on the need to take measures aimed at preventing such harm in the future.

Let us recall that according to Law N 258-FZ, experimental legal regimes can be established in the following areas:

1. Medical activities, including the use of telemedicine technologies and technologies for collecting and processing health information and diagnoses of citizens, pharmaceutical activities;
2. Design, production, and operation of vehicles, including highly automated vehicles and unmanned aerial vehicles, certification of their operators, provision of transportation and logistics services, and organization of transportation services;
3. Agriculture;
4. Financial market;
5. Distance selling of goods, works, services;
6. Architectural and construction design, capital repair, reconstruction, demolition of capital construction projects, operation of buildings, structures;
7. Provision of state and municipal services and implementation of state control (supervision) and municipal control, execution of other powers and functions by state bodies and local self-government bodies;
8. Industrial production (industry).

In addition, the Government of the Russian Federation has the right to establish other directions for the development, testing and implementation of digital innovations.

It is worth noting that the approach to AI regulation through external control over the use of technologies is also used in the European Union, including requirements for the preparation of detailed technical documentation, reporting to regulatory authorities, and the application of penalties for non-compliance with regulatory requirements (see the article "European Parliament Agrees on Restrictions for Artificial Intelligence" for more details).

• The technology community has adopted a Declaration on Responsible AI

In light of the previous news, it is interesting to note that not only the state is concerned about reducing the risks from the use of AI in Russia. The business community also takes an active position in this issue.

So, back in November 2019, leading technology companies in Russia have created an Alliance in the field of artificial intelligence for the joint development of competencies and technologies based on AI, their speedy implementation in scientific, educational and practical business areas.

In particular, this Alliance has developed a national Code in the field of artificial intelligence, which established general ethical principles and standards of behavior for participants in the domestic field of application of AI. One of the key objectives of the Code is to reduce the risks of unethical use of AI that violates human rights and interests.

Currently, more than 380 participants have already joined the national Code of Ethics for Artificial Intelligence, including 27 foreign ones from 20 countries.

In March of this year, the Alliance prepared a Declaration on the responsible development and use of services based on generative AI - a set of ethical principles, recommendations and standards of conduct in the relevant field. Sberbank, Yandex, MTS AI, leading universities and scientific organizations of the country (in particular, the Higher School of Economics, Skoltech, MIPT) signed the declaration and joined it. As noted on the Alliance website, the signing of the Declaration helps create conditions for self-regulation of the development of solutions based on generative artificial intelligence.

Abroad

• China recognized as World Leader in the use of Generative AI

In July of this year, the international news agency Reuters published information that China is currently significantly ahead in the use of AI technologies other countries, including the USA.

These results follow from an analytical survey conducted by SAS and Coleman Parkes Research across six industries: banking, insurance, healthcare, telecommunications, manufacturing, retail and energy.

China's leadership in the use of AI is also confirmed by data from the World Intellectual Property Organization. According to the latest WIPO report on the Patent Landscape on Generative AI, China has received the largest number of generative AI patents each year since 2017, significantly ahead of all other countries combined, including the United States, the Republic of Korea, Japan and India.

The WIPO report also published the main areas of application of generative AI patents, which include the following areas of activity:

1. Software;
2. Biomedical Sciences;
3. Document flow;
4. Business solutions;
5. Industry and production;
6. Transport sector;
7. Safety;
8. Telecommunications

It is interesting to note that some of the areas of activity indicated in the WIPO report are included in the list of areas in respect of which experimental legal regimes can be established in Russia under Law No. 258-FZ.

• Hong Kong Seeks to Amend Copyright Law to Stimulate AI Development

While debates in the United States rage on between copyright holders and IT companies over the use of copyrighted materials as "input data" for training generative neural networks, Hong Kong is moving to modernize its copyright legislation.

According to information posted on the Hong Kong Intellectual Property Department’s website, public consultations on improving the Copyright Ordinance to support the development of artificial intelligence technologies have been ongoing since July 8.

Specifically, there is discussion about allowing the use of copyrighted works for data analysis and AI model training without the consent of copyright holders. These public discussions will be conducted over the next two months.

It is interesting to note that in 2023, Japanese officials had already made statements about the need for free use of copyrighted materials (public data) for training generative neural networks. However, no further progress was made beyond these statements (see the article "On the Problem of Training Generative Neural Networks on Copyrighted Objects" for more details).