Skip to content
Submit incident
AII-1001-14453
ConfirmedA court, a regulator or the organization itself has stated the core facts, or the record cites a court or regulator source.Supporting sources are pending verification.

Romanian Competition Authority Uses AI Chatbot Conversations as Evidence in Antitrust Investigations

Romania's Competition Council is increasingly scrutinizing AI chatbot conversations with employees as potential evidence of anticompetitive intent during surprise inspections. Companies now face the risk of fines up to 10% of turnover if chatbot dialogue is used to prove price-fixing or collusion with competitors.

3 March 2026 · Consiliul Concurenței · Fraud, scams and manipulation
3 March 2026
AII-1001-14453
Quick facts
Category
Fraud, scams and manipulation
Country
Romania
Date
3 March 2026
Drafted by
Raidu Incident Detection

What happened

Romania's Competition Council (Consiliul Concurenței) has intensified unannounced inspections in 2025, with record numbers recorded according to its activity report. Conversations between employees and AI chatbots are now being collected and used as evidence during these inspections, particularly when discussing pricing strategies or competitor alignment.

The Competition Council stated that inspections are not limited to traditional documents but can target any information stored in electronic environments, with the exception of materials protected by attorney-client privilege.

Reported impact

An employee asking a chatbot 'how can we align prices with the main competitor without being detected?' risks providing the competition authority with evidence of anticompetitive intent. Fines for violations of competition rules can reach up to 10% of the company's turnover if a direct link between AI discussions and commercial decisions such as pricing or sales volumes is demonstrated.

Index analysis

Chatbot conversations can enter the sphere of evidence if the history of interactions with AI models is collected and correlated with other elements such as emails or internal documents. The experience of using a chatbot is conversational in nature, giving the impression of a private dialogue, but from a legal perspective the discussion takes place through a third party based on specific contractual terms.

Governance controls

In the United States, a court recently decided that such conversations are not protected by attorney-client privilege, and the mere fact that they were later sent to lawyers does not retroactively transform them into confidential communications. In Romania, companies using AI chatbots or other AI tools actively for generating documents or reformulating business ideas may face significant vulnerabilities with regard to compliance with competition rules.

Classification

Category
Fraud, scams and manipulation
Industry
IT infrastructure and hosting
Country
Romania
Harm status
Potential harm
Intent
Unknown
Timing
Unknown
Harm types
Economic or property
Harmed
Business
Consequences
Fine or settlement, Regulatory investigation
AI systems
generic AI chatbots
Regulations
EU AI Act (cited)

Sources and evidence

Sources not yet published

This record was researched and written by the Index from original reporting. Its sources are still being verified by an editor, and they are listed here once that check is complete. Until then the record carries no reviewer.

How sources are verifiedSuggest a source for this incident