ChatGPT Search by OpenAI that enables users to access real-time web information during AI conversations, has experienced significant user growth across Europe.
Highlights
According to a filing by OpenAI Ireland Limited, the tool averaged 41.3 million monthly active recipients over the six months ending March 31, 2025.
This marks a near fourfold increase from 11.2 million users reported in the previous six-month period ending October 31, 2024.
The surge in adoption brings ChatGPT Search close to the threshold of being designated a “Very Large Online Platform” (VLOP) under the European Union’s Digital Services Act (DSA), a regulatory classification that applies to services with at least 45 million average monthly users within the EU.
Approaching DSA Compliance Threshold
If ChatGPT Search crosses the 45 million user benchmark, OpenAI would be required to meet a range of regulatory obligations under the DSA.
These include offering users the ability to opt out of personalized recommendations and profiling, sharing select internal data with researchers and authorities, and undergoing independent audits to assess transparency and safety standards.
Failure to comply with the DSA can result in significant penalties, including fines of up to 6% of a company’s global annual revenue and potential temporary service suspension in the EU in cases of repeated violations.
While OpenAI has not publicly outlined its strategy for potential compliance with VLOP requirements, the rapid expansion of ChatGPT Search reflects its growing presence in the AI-assisted search space—albeit still far behind dominant players like Google, which processes vastly more search queries daily.
Search Engine Accuracy and Attribution Issues
Despite its growth, ChatGPT Search has faced scrutiny over the reliability of its responses. Independent research, including a study by Columbia University’s Tow Center for Digital Journalism, found that the tool struggled with source attribution.
Of 200 queries tested, 153 were found to contain incorrect attributions, resulting in a reported error rate of 76.5%.
Other analyses have raised concerns about how the AI handles licensed content, particularly in terms of presenting accurate summaries and maintaining proper attribution, even when sourcing from verified publishers.
OpenAI’s Response and Ongoing Improvements
OpenAI has acknowledged the challenges associated with content attribution and stated that it is working to improve the system’s accuracy.
The company has emphasized its aim to support publishers by helping users discover quality content through clear citations, quotes, and direct links. Updates to the feature are reportedly focused on refining how content is summarized and credited to original sources.
Regulatory Implications
The possible classification of ChatGPT Search as a VLOP highlights broader regulatory conversations around emerging AI tools.
The EU’s DSA is one of the first major frameworks attempting to regulate algorithmic content distribution, platform accountability, and user rights in the context of generative AI and conversational interfaces.
While conversational search powered by AI continues to evolve, its increased usage suggests a shifting landscape in how users seek and interact with online information.
Tools like ChatGPT Search offer a blend of real-time data and natural language processing, introducing both new possibilities and challenges for digital governance.
As user adoption accelerates and oversight becomes more defined, ChatGPT Search is likely to serve as a key reference point in ongoing discussions about regulating large-scale AI applications in Europe and beyond.
