AI Strategies in Poland

Yes. The Polish Ministry of Digital Affairs (the “Ministry”) is in charge of the initiative to introduce a legal regulatory framework for AI. 
The Ministry cooperates with the Ministry of Entrepreneurship and Technology, the Ministry of Science and Higher Education, and the Ministry of Investment and Economic Development

N/A

No draft regulations have been prepared to date. At the moment, major deliverables of the initiative include:

• AI Strategy Paper (the “AI Strategy”), issued on 9 November 2018 by the Ministry. The AI Strategy addresses both the commercial and legal consequences of AI operations;
• Artificial Intelligence Development Policy in Poland for 2019–2027 (the “AI Policy”); issued by the Ministry on 21 August 2019, together with the other Ministries. The AI Policy is a development which implements recommendations from the AI Strategy in a greater detail.

The adopted definition of AI follows the recommendation of the EU High-Level Expert Group (HLEG) on AI (EC) expressed in the document Definition of Artificial Intelligence, as well as the recommendation of AIGO group (OECD). 

The Ministry opts against granting AI a legal personality or limited capacity to perform legal acts as it is contrary to human centrism (the higher status of humans over machines) and it is not proportionate to AI’s current state of development. 

The prevailing concept is that AI as a technological solution without a legal capacity cannot be held liable for damage. For the sake of cohesion, the Ministry calls for international rules to be developed for liability for damage caused by AI.

For this reason, several concepts are currently being considered as to who should be held liable, with some preference for strict liability for damage caused by AI. An acceptable compromise could be a liability regime similar to liability for products or a stand-alone liability regime for damage caused by AI.

Temporarily, liability for dangerous products, adopted appropriately, should be relied on.

So far there are no detailed concepts on the safety compliance of AI systems developed by the Ministry.

NASK has been appointed as a cybersecurity centre for: (i) monitoring, researching and responding to threats to AI systems; and (ii) the standardisation and certification of AI systems’ cybersecurity procedures. 

The Ministry claims that the transparency of AI Systems should ensure: 1) that the user is aware that he/she is interacting with AI; 2) access to the underlying algorithms and understanding of how AI works; 3) contact with a human (if necessary) and ability to check or correct decisions taken by AI. 

The Ministry also states the importance of determining which decision-making processes can/cannot be entrusted to AI and enable access to data on all the operations and decision-making process (the “black box”).

The Ministry strongly opposes an absolute right of “ownership” of non-personal (machine) data. Such data is also not a work and, as such, is not subject to protection under intellectual property law. 

Instead of a separate “data ownership right”, a legal framework regarding access to such data should be developed, taking from the GDPR principle of data portability. Differences in access should take into account sectoral priorities and needs. 

The Ministry is against granting AI copyrights. For clarity, the Ministry recommends explicitly indicating that the author of a work can only be a human under Polish copyright law.

Nonetheless, the Ministry reserves the possibility of a need to provide legal protection for works created by AI in the future. 

The Ministry identifies a need for legislative interventions regarding the development of AI, such as the creation of organisations to certify AI in terms of transparency, accountability and impartiality, following compliance rules for Trustworthy AI, as described in the Ethics Guidelines for Trustworthy Artificial Intelligence provided on 8 April 2019 by the EU High-Level Expert Group.

The Ministry supports the need for human supervision over the functioning of AI. 

Please see the sections regarding the legal personality of AI and liability for damage caused by AI for more details.

The Ministry does not propose concrete solutions, but acknowledges competition law issues. These include: 1) ensuring transparent use of AI; 2) ensuring the monitoring of the market, particularly regarding the algorithms used by AI to determine prices or manage access to resources.

The Ministry highlighted the need to: (i) fully and truly inform consumers about the scope, mechanisms and entities using consumer data and their impact on the legal relationship; and (ii) specify the position of the “supplier” of AI systems, including the supplier’s duties and responsibilities.

The Ministry proposed introducing the AI Observatory for Changes in the Labour Market, a study of the adequacy of legislative intervention regarding, e.g.: (i) the taxation of the work of robots; (ii) charges to employers for the use of AI solutions leading to the loss or reduction of workplaces; and (iii) the creation of a fund with a mandatory contribution paid by employers using AI solutions, with the resources collected to support the unemployed.

The Ministry noted that legal provisions regarding outsourcing services may apply to financial institutions’ use of the services of AI providers, which may discourage AI providers from entering into agreements with financial institutions. Hence, the Ministry calls for legislative initiatives to remove the potential barriers.

No solutions were proposed in this regard. 

The Ministry pointed out a need to analyse the demand for legislative work on defining the basic principles of the automation of administrative activities.