The new technology sector has responded positively to the ongoing work on the draft law on artificial intelligence (AI) systems. However, it has also pointed out several inconsistencies and potential threats to AI development in Poland under the new regulations proposed by the Ministry of Digital Affairs. According to industry representatives, these provisions require amendments. This is the position expressed by the Digital Poland Association in a letter sent to the Minister of Digital Affairs, Krzysztof Gawkowski, on behalf of digital entrepreneurs. The organization’s concerns primarily relate to the law’s consistency with EU regulations, the mechanism for imposing potential sanctions on technology providers, and the supervision and control procedures for entities offering AI-based products or services.
Key Concerns Raised in Public Consultation
The letter from the Digital Poland Association was submitted as part of the second round of consultations on the AI systems bill. According to Michał Kanownik, President of the Digital Poland Association, the proposed law is crucial for the proper development of AI in the country.
“We appreciate the Ministry of Digital Affairs for considering the comments and suggestions submitted during the first round of consultations. The current proposal aims to protect citizens’ rights while enabling technological innovation, which is key to the Polish economy. No other technology today offers as much potential as artificial intelligence,” emphasized Kanownik.
At the same time, Digital Poland experts have proposed several amendments, arguing that they are necessary to eliminate regulatory uncertainty and ensure AI development in Poland is not hindered.
Unclear Provisions on AI Supplier Oversight and Data Privacy Risks
A major concern raised by experts in their letter to the ministry is the oversight and control mechanism for entities supplying AI-based solutions. Digital Poland points out that the bill does not define the qualifications of inspectors conducting AI system audits, nor does it impose obligations regarding confidentiality of information obtained during inspections. This creates the risk that an inspector could be a competitor of the audited entity or have a personal interest in issuing an unfavorable inspection report.
Additionally, the organization highlighted data privacy concerns. The current provisions are so broadly formulated that designated inspectors could request access to customer content stored in cloud accounts, for example. The industry warns that this not only threatens personal data privacy but also contradicts the principles of the EU AI Act. Digital Poland insists that the law should allow AI providers to classify certain information as trade secrets, similar to how the Polish Office of Competition and Consumer Protection (UOKiK) handles inspections.
The organization also advocates for stricter safeguards regarding the disclosure of AI-related information to authorities such as the police. It proposes requiring prior court or prosecutor approval for such disclosures.
“AI-related data often consists of highly specialized, valuable business and market information. Given the potential consequences of unauthorized disclosure, mechanisms must be established to prevent unjustified sharing of such data,” the letter states.
Financial Penalties Could Stifle AI Development
Industry experts also raised concerns about potential criminal sanctions against AI technology providers. Digital Poland particularly opposes a provision that grants the AI Development and Security Commission the authority to order an AI system operator to cease using a solution or withdraw it from the market. The organization argues that this provision is overly vague, allowing for far-reaching decisions to be enforced immediately without a real possibility of appeal in court. The lack of clear decision-making criteria contradicts legal certainty and predictability, according to industry representatives.
Another controversial aspect of the bill is its provision allowing the Commission to finance its operations through fines imposed on AI providers. Digital Poland argues that this could incentivize the imposition of excessive penalties to improve the Commission’s financial standing.
“We believe this approach could be highly detrimental to AI development in Poland. A government agency should not be systematically incentivized to impose the highest possible number of financial penalties on AI providers,” commented Kanownik.
Moreover, the association warns that such regulations could undermine AI-friendly initiatives like regulatory sandboxes, which allow businesses to test AI solutions in a controlled environment.
Call for a More Balanced and Transparent AI Commission
The Digital Poland Association also proposes expanding the composition of the AI Development and Security Commission to include representatives from academia and business organizations. The industry believes that this would ensure a more transparent and balanced approach to AI development.
“The government should ensure that the AI law aligns with both the letter and spirit of the EU AI Act. If the concerns and comments we have outlined are not addressed, they could lead to significant regulatory uncertainty and ultimately hinder AI development in Poland,” Kanownik warned.
