Poland’s Government Is Leveraging Machine Learning for Public Good

Table of Contents

In recent years, Poland’s government has embraced machine learning (ML) not merely as a technological novelty but as a core pillar of its national development strategy. By combining substantial public funding, forward-looking legislation, and targeted pilot projects, Warsaw aims to transform everything from public administration efficiency to national security. This article provides an in-depth exploration of how Polish authorities are harnessing ML—drawing on official initiatives, expert commentary, and public-sector pilots—to benefit citizens across the entire country.

1. Setting the Stage: National AI Strategy and Regulatory Foundations

Poland’s journey towards systematic ML integration began with high-level policy measures in 2024. On 21 January 2024, the Minister of Digital Affairs announced the creation of a dedicated AI Development Fund, earmarking resources to support home-grown AI start-ups and research endeavours. Simultaneously, the government convened the “PL/AI – Artificial Intelligence for Poland” advisory board, composed of leading scientists, entrepreneurs, and technologists charged with crafting guidelines for responsible AI deployment.

To give legal force to these ambitions, the Ministry of Digital Affairs submitted a draft Act on Artificial Intelligence Systems in October 2024. Following a six-week public consultation, a revised version was published on 11 February 2025, proposing the establishment of regulatory sandboxes and an AI supervisory authority under the Committee of Ministers for Digital Affairs. Alongside legislative efforts, the Polish Parliament set up a Standing Subcommittee on AI and Transparency of Algorithms to assess societal impacts and risks, ensuring that algorithmic decision-making remains fair and explainable.

2. Financing Innovation: The 1 Billion Złoty AI Development Plan

Ambitious policy declarations are backed by real money. In November 2024, the government unveiled a 1 billion złoty (approximately US $240 million) AI Development Plan to boost economic competitiveness and defence readiness. Administered through both the AI Development Fund and the National Centre for Research and Development (NCBR), these investments finance grants, technology incubators, and “IDEAS NCBR” R&D programmes aimed at accelerating ML-driven solutions across sectors.

By aligning grant programmes with industry needs—including manufacturing, health care, and logistics—Poland hopes to foster a robust ecosystem of AI enterprises. An Ernst & Young study from 2024 found that 21 percent of medium and large Polish manufacturers already employ AI in production, indicating fertile ground for further ML innovation.

3. From Models to Ministries: PLLuM and Pilot Deployments

Translating policy into practice, the Ministry of Digital Affairs has partnered with the HIVE consortium—led by Dr Agnieszka Karlińska—to pilot a suite of domestic AI models known as “PLLuM” (Polish Language Models) in 2025. The consortium, which includes the Central IT Centre (COI) and Cyfronet AGH (home to Poland’s most powerful supercomputer), is fine-tuning 18 ML models and chatbots for government use.

Key applications include:

• Administrative Assistants: Chatbots that streamline routine inquiries for civil servants, reducing bureaucratic overhead and standardising responses.

• mObywatel Integration: Embedding ML-powered virtual assistants into Poland’s flagship mObywatel app to help citizens navigate e-services such as registrations, benefits applications, and status tracking.

These pilot implementations—scheduled to expand to additional ministries by year-end—will test real-world efficacy, user satisfaction, and security hardening before a nationwide rollout.

4. Modernising Public Services: AI in Everyday Administration

Public sentiment towards ML in administration is cautiously optimistic. A 2024 survey by the Polish Economic Institute revealed that 60.4 percent of Poles expect broader AI adoption in public services, particularly for automating processes and improving accessibility. Existing platforms such as ePIT (electronic tax filing) and mObywatel already boast high satisfaction—92.5 percent of users agree they ease official procedures—creating a solid digital foundation for AI enhancements.

Practical applications under exploration include:

• Document Processing: Using natural language processing (NLP) to extract and verify data from scanned forms, speeding up applications for permits, licences, and social benefits.

• Predictive Case Routing: ML algorithms that assess case complexity and route citizen inquiries to specialised teams, reducing wait times and misclassification.

• Proactive Notifications: Intelligent reminders for upcoming deadlines (tax filings, licence renewals) delivered via SMS or the mObywatel app, driving compliance and reducing black-market fines.

By focusing ML efforts on high-volume, repetitive tasks, the government aims to free human agents for nuanced, value-added interactions.

5. Safeguarding Society: AI Ethics, Privacy, and Security

Never far from public debate, ethical and privacy concerns shape Poland’s ML agenda. During the AI Act consultations, data protection surfaced as a key issue: 32 percent of respondents feared AI could breach personal data or reduce transparency. To address these worries, the draft legislation mandates:

• Explainability Requirements: Crucial decisions (e.g., welfare eligibility) must be accompanied by human-readable explanations of algorithmic reasoning.

• Data Minimisation: Models deployed in public administration must use only the minimum data necessary, with stringent anonymisation protocols.

• Regulatory Sandboxes: Controlled environments where new AI systems can be tested for safety and bias under official supervision before public release.

These frameworks aim to build public trust while preventing misuse—especially vital as ML systems permeate sensitive domains such as health care, social services, and immigration.

6. Beyond Civvy Street: Defence and National Security Applications

Machine learning’s appeal extends to national security. In August 2024, Poland’s Ministry of National Defence unveiled a 15-year Military AI Strategy (2024–2039), setting out a phased roadmap for responsible AI deployment in line with NATO standards. Emphasis falls on human-in-the-loop oversight, robust testing environments, and collaborative R&D with allied partners. Key objectives include:

1. Establishing an AI Implementation Centre for strategic coordination and training.

2. Modernising Data Infrastructure to feed ML models for surveillance, logistics optimisation, and predictive maintenance.

3. Fostering International Collaboration—notably via shared exercises with NATO allies and information-sharing with the US Department of Homeland Security (DHS) under a 2024 MOU.

These defence-oriented AI projects, while less visible to the public, strengthen Poland’s resilience against emerging threats and cyber-espionage.

7. Cultivating a National AI Ecosystem: Academia, Industry, and Collaboration

A thriving ML ecosystem demands more than government programmes; it requires academia-industry linkages and confidence from private investors. Key players include:

• National Centre for Research and Development (NCBR): Offers large grants through its IDEAS programme to universities and start-ups for proof-of-concept AI projects.

• Polish Economic Institute (PIE): Works with ministries to gauge public attitudes and identify service gaps ripe for AI solutions.

• Innovative Start-ups: Firms such as Samurai Labs, DeepSense.ai, and Infermedica are leveraging ML in sectors ranging from medical diagnostics to predictive manufacturing maintenance.

Additionally, Poland’s National AI Days conferences foster networking, while university-led HIVE and EU Horizon Europe grants nurture cross-border research.

8. Challenges Ahead: Skills, Adoption, and Regional Equity

Despite measurable progress, several hurdles remain:

• Talent Shortage: Poland must scale up AI education at universities and vocational schools to meet growing demand for data scientists and ML engineers.

• Digital Divide: Ensuring equitable ML-powered services for rural areas—where internet penetration and digital literacy lag behind urban centres—poses an ongoing challenge.

• Legacy Systems: Integrating AI with entrenched IT infrastructures in local governments requires careful change management to avoid disruption.

Policymakers are responding with training subsidies, regional digital hubs, and phased technology roll-outs to bridge these gaps.

9. Looking Forward: A Machine-Learning-Powered Welfare State?

By 2026, Poland aims to transition from experimental pilots to mainstream ML integration across key ministries. Priorities include:

• Health Care: Predictive analytics for hospital resource allocation and early detection of disease outbreaks.

• Education: Adaptive learning platforms that personalise curricula based on student performance data.

• Transportation: Real-time traffic management and predictive maintenance of national rail and highway networks.

Ongoing collaboration with the European Union’s AI Act framework and Horizon Europe research programmes will further align Poland’s efforts with continental best practices.

Conclusion

Poland’s governmental embrace of machine learning marries pragmatic public-sector efficiency gains with a broader vision of digital sovereignty. Through generous funding, robust legal safeguards, and high-profile pilot projects—such as the PLLuM model deployments and AI regulatory sandboxes—the country is crafting a blueprint for responsible, citizen-centric AI adoption. As Poles increasingly expect ML-driven convenience in services ranging from tax filings to healthcare, Warsaw’s balanced approach—anchored in transparency, ethics, and international cooperation—positions Poland as a rising star in Europe’s AI landscape.

By fostering a vibrant ecosystem that spans ministries, academia, and private innovators, and by proactively addressing challenges of equity and skill development, the Polish government is laying the foundations for a machine-learning-powered welfare state—one that not only automates tasks but also enriches democratic governance and national resilience.