Enforcing the Law on the Fly: The New Architecture of Digital Market Supervision in Poland

2026-06-09 • Mariusz Jażdżyk

For years, we tried to control algorithms with paper. It never stood a chance. The government has just adopted legislative packages implementing the EU Digital Services Act (DSA). This is a rare moment when the state stops merely observing the market and begins organizing a solid front.

The architecture of accountability has been distributed. The Office of Electronic Communications (UKE) assumes coordination, the National Broadcasting Council (KRRiT) oversees video, and UOKiK verifies interface manipulations. This is a fundamental remodeling of relations and an attempt to level the asymmetry between the state and corporations. Changing regulations, however, is merely a declaration of intent.

The Algorithmic Battlefield

The real battle for digital markets will be decided in the execution layer of the code. Platforms optimize conversion in milliseconds, utilizing dynamic A/B tests and asynchronous scripts. Manually investigating these phenomena is technically impossible. It requires infrastructure capable of analyzing millions of variables in real-time.

Law without enforcement tools is merely a suggestion. Instead of evaluating static mock-ups, supervision must verify the actual behavior of systems on the fly. Autonomous swarms of AI agents simulate purchasing paths. They wait for promotional timers to expire and listen to API queries in search of hidden costs.

Inspector-in-the-Loop

Deploying software alone does not solve the problem. Technology requires trust, and trust requires a deep understanding of context. The answer is an operational model we can call Inspector-in-the-loop. The official ceases to be an administrator clicking on screens.

The expert defines the rigorous parameters of the task, oversees the runs, and corrects errors. In this way, they transfer their unique, substantive domain knowledge directly to the algorithms. Ultimately, they become the manager of a powerful infrastructure. They gain scale, becoming a real threat to Big Techs abusing their position.

This is an extremely difficult transformation for public administration. However, it may bring an unexpected result. The mission of technologically leveling the playing field with global giants has the potential to attract young, capable talents to government offices. Administration can once again become a place where one genuinely and tangibly influences market reality.

Enterprise-Grade Administration

Recent deployments demonstrate that the Polish administration is ready for this leap. Poland is slowly becoming a digitalization leader in areas critical to citizens. On one hand, there is the medical sector, where advanced AI platforms are becoming available to hospitals, optimizing diagnostic processes. On the other hand, there is consumer protection, where UOKiK tests and implements algorithms for automatically detecting dark patterns and mapping prohibited clauses in contracts.

For e-commerce boards, this is a clear signal. Compliance with consumer law is no longer a matter of legal interpretation. It becomes a hard engineering requirement, embedded directly in system architectures. Manipulations in the code cease to be invisible.

Shift-Left Compliance as the New Standard

When supervisory bodies acquire Enterprise-grade technology, hiding asymmetry in interfaces loses any economic sense. This opens the way for a positive evolution of the entire market. Instead of waiting for inspections and fines for DSA violations, aware e-commerce organizations now have the opportunity to shift legal verification directly to testing environments (Shift-Left Compliance).

Integrating audits at an early stage of code development will make cancellation or return processes as simple as the purchase itself. Fairness by Design then ceases to be merely a way to avoid penalties. It becomes a natural standard that builds long-term trust and ultimately guarantees consumers full autonomy of choice. Such a move is in the interest of both parties.

When machines begin verifying code on a massive scale, humans will regain the time to judge intentions. When the market optimizes via algorithms, supervision must become an algorithm.