The recent "Dialogues in Competition Law and Economics" conference, held under the auspices of the Hellenic Competition Commission, offered inspiration for reflecting on how digital transformation is reshaping competition enforcement across Europe. An area of great interest is the growing reliance on pricing algorithms in digital markets. 

Although not directly discussed during the conference, it perfectly illustrates the tension between technological efficiency and market fairness, as pricing algorithms can enhance responsiveness , yet they may also facilitate collusion, enable discriminatory pricing, or reinforce exclusionary practices. Under the EU’s results-based enforcement model, undertakings cannot rely on intent or good faith to avoid liability. 

Conduct producing anti-competitive effects remains sanctionable — as shown in Eturas (CJEU, C-74/14), where travel agencies using a common platform faced liability due to an algorithmic cap on discounts, or in the UK CMA’s Amazon poster seller case, where competitors’ reliance on identical pricing software amounted to a cartel. At the same time, the newly adopted Artificial Intelligence Act (Regulation (EU) 2024/1689) introduces a preventive, governance-based layer. 

While pricing algorithms are not listed among the prohibited practices, they may qualify as high-risk systems when they significantly affect consumers’ economic interests. The Act establishes obligations of risk management, transparency, data governance, and human oversight, embedding responsibility directly into design and deployment. Competition law and the AI Act thus operate in parallel: competition law intervenes ex post to sanction anti-competitive outcomes, while the AI Act imposes ex ante duties of care, requiring undertakings to anticipate and manage risks before deployment. This dual framework means that firms may face exposure on both fronts — for the collusive results of algorithmic behavior and for inadequate AI governance. Absolute immunity is not possible , yet structured compliance can mitigate liability. 

What truly matters is cultivating a culture where algorithms are designed with internal safeguards against collusion, ensuring qualified human oversight, maintaining transparency when AI influences prices, and documenting every step as evidence of responsible governance. For online marketplaces the message is clear: compliance by design, human oversight, and transparency may not be guarantees against enforcement but they are the most reliable shield in a results-based regulatory environment that now demands accountability for both design and outcomes. 

How should competition authorities and businesses adapt to AI-driven pricing?