AI Strategies in Hungary

The Hungarian AI Coalition (MI Koalíció) was formally created on 9 October 2018. It is a partnership between state agencies, leading IT businesses, and universities, and works towards implementing a comprehensive AI strategy for the future, which industry leaders hope will establish this country as an AI innovator. In addition to formulating a national strategy, the AI Coalition is studying the social and economic effects of AI on society. As further proof of Hungary’s commitment to AI research, in April 2018 the Hungarian government signed the EU Declaration on Artificial Intelligence, a cooperation agreement among 24 member states for the research, implementation, and regulation of AI with the ultimate aim of increasing the EU’s global competitiveness in this sector.

In May 2020, the Hungarian government adopted a new comprehensive AI Strategy (Magyarország Mesterséges Intelligencia Stratégiája) that sets ambitious goals and measures for the development and use of AI in Hungary through to 2030.

The focus areas of the AI Strategy identify sector-specific development goals designed to help Hungary occupy an elevated position in the global value chain. The English version of the AI Strategy can be found at the following link: http://https//ai-hungary.com/en.

The AI Strategy proposes the creation of interdependent measures, such as foundation pillars, focus areas and transformative programmes. The foundation pillars help prepare society to manage the inevitable changes resulting from AI and to exploit effectively and fully the technology’s advantages as follows:

1. The aim of the first foundation pillar is setting the data economy in motion by introducing a data-market platform and making public data available in line with EU legislation. The aim of the data-market platform is to establish and raise awareness of the economic value of commercially viable data, and to encourage its secondary use.
2. The foundation pillar for research, development and innovation will lead to the creation of the National Artificial Intelligence Laboratory (AI LAB - Mesterséges Intelligencia Nemzeti Laboratórium), which will act as a coordinating body between the research institutes, the market, the AI research scene and the international research community. Hungary is expected to use the AI LAB as a National Laboratory for pursuing global research projects. To facilitate this research, the Laboratory will act as a coordinating body between research institutes, the market, the AI research scene and the international research community. The Lab will provide funding for flagship basic and applied research projects (e.g. publications, patents, emerging industrial relations, technology transfer) and strengthen the synergy between those actors representing both industry and international projects.
3. The foundation pillar for education, competence development and societal preparedness will lead to the creation of an AI Innovation Centre (MI Innovációs Központ) for collecting and disseminating information, event organisation and social dialogue, the development of profession-specific training techniques and their application to the education system.

Another foundation pillar is the regulatory and ethical framework, which will require the creation of a general regulatory environment for data assets.

N/A

No draft regulations have yet been proposed. However, the Regulatory Working Group of the AI Coalition discussed the draft wording of the regulatory aspects of the AI Strategy. The AI Strategy does not contain harsh statements and the regulation can address ethics as well as sectoral and facilitating solutions.

The AI Strategy describes AI as “a piece of software capable of mapping parts of human intelligence, and supporting or autonomously performing processes of sensing, interpreting, decision making and action”. In the AI Strategy, AI is understood to mean “narrow” AI, i.e. systems only capable of mapping specific areas of human intelligence. Research into “general” AI—capable of mapping all human intelligence—remains underdeveloped and uncertain, and therefore does not apply within the context of the AI Strategy.

Otherwise, neither the AI Strategy nor the Regulatory Working Group of the AI Coalition provide for the establishment of a more specific definition of AI.

The AI Strategy proposes the creation of an AI regulatory environment which will include the concept of “technological entities”. Otherwise, neither the AI Strategy nor the Regulatory Working Group of the AI Coalition provide for a specific concept of the legal personality or capacity of AI yet.

The AI Strategy proposes developing an AI regulatory environment (comprising rules on registration, technological entities, responsibility/ liability and industry regulations). The aim is to explore legal constraints on and the regulatory needs of AI development and to make proposals regarding changes to be made to the general regulatory environment, along with improvements to the sector-specific regulatory environment in order to facilitate AI development. This will include the separation of liabilities and provision of continuous support for rules and developments.

Otherwise, neither the AI Strategy nor the Regulatory Working Group of the AI Coalition provides for any specific concepts regarding the liability of AI systems.

Apart from setting out the general concept of regulation liability in the context of AI (as outlined above), neither the AI Coalition nor the Regulatory Working Group of the AI Coalition provides for any concepts regarding the safety compliance of AI systems.

One of the main objectives of the AI Strategy is that the measures defined mainly serve to facilitate and encourage data processing while fully respecting the custodians’ rights set forth in law, and the limits specified by domestic and international legislation for the use and processing of the data, with special attention to the protection of personal data, including healthcare-related and other sensitive data.

The AI Strategy aims to establish and raise awareness of the economic value of commercially viable data, and to encourage their secondary use to take advantage of their potential as a resource of the future. The instrument for this is the introduction of a Data Market Platform (Adatpiac Platform) in three stages:

In the first stage, a marketplace exercising brokerage function will be established, which will provide a space for the transmission of non-personal data and the related technologies. After that, the aim is to further develop a marketplace that is capable of more than brokerage functions and will be able to provide data storage as well as data quality based on standards and certificates. The second stage will involve the creation of commercial opportunities for data assets including personal data, and preparing the introduction of individuals as actors into the data market (linked to the development of a related mobile app). The third stage will include the creation of the “One Stop Shop” platform, which will operate as a platform serving data custodians, as well as the researchers and developers of the application, with broad-based functionality, by integrating the data assets of the data markets and data spaces of the EU, and by the active use of data spaces of the EU in the course of using Hungarian data.

According to the AI Strategy, only anonymous personal data will be included in the machine-learning models, will be an object of trade, and will be transferred to third parties. However, anonymous data will be able to be decrypted using technological and inferential equipment, depending on the context of use. To ensure secured data trade that takes user rights into full consideration and thus to ensure AI development, it will be critical to define new technological and process developments that can guarantee the unambiguity of personal data.

The purpose of the data wallet initiative (adattárca) under the AI Strategy is that both companies and private entities will be interested in participating—in a transparent way—in the data economy, providing all stakeholders with a regulated framework for maximising the benefits from the accumulation of the data assets, based on the progressive opportunities provided by the GDPR and the PSI II.

Apart from the creation of the Data Market Platform as outlined above, the AI Strategy also aims to create a “Portal for Public Data” (Közadatportál) to make available public data based on EU legislation. The aim is to create a comprehensive, responsible public data management system, which includes the sharing of non-personal data sets created in the public sector to aid economic recovery. It also aims to develop an integrated governmental approach which will be able to enforce the benefits of economic growth against the investments necessary for the open access of the data.

To connect public data with the data cycle, a new body, the National Data Assets Agency (Nemzeti Adatvagyon Ügynökség - NAVÜ) will be established, to coordinate and stimulate the strategic management of the data assets as the engine of utilising national data assets. The NAVÜ is being established as part of Neumann János Nonprofit Ltd., whose ownership rights are exercised by the Ministry of Innovation and Technology. The NAVÜ’s main task is to create technical and semantic standards for transparent traffic and secondary use of public-data assets. It acts as a coordinator in creating a comprehensive public-data platform from existing public databases that will allow for the secondary use of existing data. In addition to providing easy access to data for research and development purposes, the platform will monitor the use of public data and allow for the exploitation of economic opportunities due to the wide availability of national data assets in the EU.

For the sake of transparency, the establishment of a public data portal is required, which will give access to the data and provide a one-stop connection opportunity to access national data assets freely—based on EU legislation—and, on the other hand, make the relationship between the state and businesses (G2B), the state and clients (G2C) and states (G2G) possible.

It is a key objective to ensure market-based trading in data that has a significant multiplier effect and which is created in the public sector but not subject to Directive (EU) 2019/1024 on open data and the re-use of public sector information by connecting the Data Market Platform and the Public Data Portal.

The aim is fully integrating the secondary use of the data created in public administration with the data cycle and to establish an effective model for the management of public data. It is necessary to identify those groups of data, the commercial use of which has already started, or which may be involved in commerce, and NAVÜ and the Data Market Platform will be able to promote the effectiveness of their commerce.

The Data Asset and Regulatory Working Group of the AI Coalition held a joint workshop whose goal was to determine the definitions and basic elements of new legislation on monetising data sets.

The workshop identified four ways that data and data sets can be regulated in Hungary from a civil law aspect:

• ownership rights;
• copyright protection;
• trade secret protection;
• simple “rights”.

Finally, the Regulatory Working Group recommends a new sui generis right for the protection of data and data sets that would need to meet the following criteria:

• it must be a general solution;
• this new right must fit into existing EU definitions.

The AI Strategy includes a specific chapter on establishing a regulatory and oversight framework to ensure responsible, trustworthy and people-centred utilisation of the technology’s potentials. The key goals are the creation of an effective, efficient and supportive domestic regulatory environment for the operation of AI, together with the necessary ethical framework, and taking the relevant EU legislation into account.

This will include:

Creating a regulatory framework for data assets. A general regulatory environment needs to be put in place for data assets with functions such as supporting the AI-related use of public data assets and facilitating the process of turning data into assets (monetisation), together with the development of the relevant financial and legal regulations, taking into account the various sectors’ specificities and responsibilities in terms of data processing, as well as the relevant fundamental rights and the international framework of data regulations.

More specifically:

• creating a framework law on data assets;
• introducing a sector-specific regulatory environment to enable data to be turned into assets and used for purposes of AI;
• developing rules to govern the use of public data, along with a concept for and rules on their monetisation.

Developing an AI regulatory environment (comprising rules on registration, technological entities, responsibility/liability and industry regulations). The aim is to explore legal constraints on and the regulatory needs of AI development and to make proposals regarding changes to be made to the general regulatory environment, along with improvements to the sector-specific regulatory environment to facilitate AI development.

More specifically:

• continuous monitoring and mediation of the relevant EU rules and other regulations (including developments in legal practice and jurisprudence), participating in the preparation of legal regulations and quick and effective communication of new rules and interpretations to the Hungarian legal system;
• developing AI registers and deciding requirements to be applied in the most important areas;
• separating liabilities and the provision of continuous support for rules and developments;
• supporting the implementation of the AI Strategy in terms of statutory rules and ethics.

Creating an ethics framework for the AI industry. The aim is to create a Code of Ethics, taking international standards and domestic specificities into account, based on a broad consensus and factoring in the work and output of the European Commission, as well as the Ad-hoc Committee on Artificial Intelligence of the Council of Europe. The development of the required ethics framework will involve specifying what is meant by the term humancentric, identifying ethical AI objectives and development directions, and integrating these in a reliable and robust AI framework, as well as developing techniques for asserting and enforcing the rules of ethics.

Creating an Artificial Intelligence Regulation and Ethics Knowledge Centre (Mesterséges Intelligencia Szabályozási és Etika Tudásközpont - MISZET). The aim is to create and coordinate an extensive pool of experts to help resolve legal issues and matters of ethics relating to the regulation of artificial intelligence and the implementation of the strategy.

Although the AI Strategy mentions the development of trustworthy AI as the key direction for AI research and development, it declares that one of the barriers to the broad use of AI originates from the limitations of the current state of technology. According to the AI Strategy, in line with efforts made by the EU, Hungary will focus on eliminating possible forms of error in AI, or making it transparent and predictable (developing test environments), making model decision-making mechanisms explicit, or developing hybrid models. However, the AI Strategy does not mention human oversight as part of these efforts. The Regulatory Working Group of the AI Coalition has also not yet provided for any concepts regarding human oversight.

In September 2018, the Hungarian Competition Office (Gazdasági Versenyhivatal – GVH) declared in its Medium-term Digital Consumer Protection Strategy (Középtávú digitális fogyasztóvédelmi stratégia) that it must ensure that businesses should not use, e.g. algorithms or AI to the detriment of consumers, and consumers without digital literacy should not lag behind due to the lack of information.

Otherwise, neither the AI Strategy nor the Regulatory Working Group of the AI Coalition has provided any concepts regarding competition yet.

The AI Strategy contains an action plan for the completion of the AI Challenge in Hungary by the end of 2021: awareness raising for one million people and training for 100,000 people to complete a basic AI course.

In addition, the AI Hungary website at https://ai-hungary.com/ will contain online tech demonstrations and provide education materials and online tests.

The AI Strategy contains a specific section for the analysis of the changes in the labour market. One million citizens perform new or higher value-added AI-supported work due to a change in position or workplace. An estimated 900,000 employees will be affected by AI and automation by the middle of the 2030s. The Hungarian government aims for more employees to be able to perform work that has high added value and provides a better quality of life by 2030.

The AI Strategy envisages that retraining and upskilling will be a central issue for the labour market up to the 2030s regarding the system of skills in the course of renewal in the twenty-first century. The assessment of supply and demand in the current activity of the National Employment Service (Nemzeti Foglalkoztatási Szolgálat - NFSZ) must also be shifted to competence based recruitment, e.g. through the Hungarian adoption of the AI technologies of the ESCO system and employment services. A change of such magnitude requires coordinated preparation at a societal level, with everybody’s involvement.

In the planning, preparation and implementation of development projects, the relevant organisational units of the NFSZ need to play an important role in the appropriate forecasting of and support for labour market trends resulting from the technological transition.

The AI Strategy envisages the development of systems supporting the oversight of financial and taxation processes. This will include improving the efficiency of the processes employed by the State Audit Office, the National Tax and Customs Administration and the banking supervisory authorities by pre-screening items to be investigated and by introducing process automation (software robots, language processing).

Otherwise, the Hungarian Government and the Hungarian AI Coalition has not discussed any concepts that might affect the impact of AI on the financial sector yet.

The AI Strategy proposes creating a regulatory framework for data assets with functions such as supporting the AI related use of public data assets and facilitating the process of turning data into assets (monetisation), together with the development of the relevant financial and legal regulations, taking into account the various sectors’ specificities and responsibilities in terms of data processing, as well as the relevant fundamental rights and the international framework of data regulations. See the “Regulatory” section for more details.

Additionally, the Hungarian AI Coalition plans to support the Hungarian digital economy with the acceleration of client management for state entities. The meetup of the Hungarian AI Coalition on the AI Strategy and transformative projects (MI Meet-Up: Szólj bele a magyar mesterséges intelligencia stratégiába!) focuses on, among others, the digitalisation of administration and language tools.