1. After opening on 9 January 2026 an ex officio investigation in the conversational agents sector, the French Autorité de la concurrence (the “Authority”) has launched on 29 January 2026 a public consultation in this regard (the “Consultation”), marking another step in the Authority’s scrutiny of the generative Artificial Intelligence (“AI”) value chain.[1]
2. Stakeholders still have a bit less than two weeks to respond to the Consultation as their submissions should in principle be filed before 6 March 2026.
3. The Authority defines conversational agents as a “generative AI-based tools that interact with users in real time in natural language, capable of responding to queries or assisting users with tasks in an automated manner.”[2]
4. This Consultation builds on the Authority’s previous studies on AI, including its 2024 opinion on generative AI foundation models[3] and its 2025 opinion on the environmental impact of AI,[4] which together laid the groundwork for understanding both the upstream and sustainability aspects of the AI ecosystem.
5. While the geographic scope of the Consultation is primarily France, the questions raised have clear EU-wide implications due to the cross-border nature of digital markets, creating an incentive for all EU stakeholders to respond to the Consultation.
6. The Consultation therefore offers all the stakeholders such as companies offering generative AI services, cloud service providers, e-commerce platforms and e-retailers, advertisers, publishers and conversational agents’ users a concrete opportunity to shape how the Authority (and by capillarity potentially other competition authorities) will make its competition assessment of the conversational agents sector in the coming years.
7. Responding to the Consultation will contribute to shaping the market boundaries in the sector and explaining its competitive dynamics (Section I). It will also allow the Authority to understand better the emerging commercial uses of conversational agents. (Section II).
- Understanding the conversational agents sector
8. To ensure fair competition in the sector, it is critical that the Authority has a clear understanding of the structure of the market and the interactions between market players (A). Stakeholders’ contributions will also play a decisive role in shaping how the Authority assesses behaviour of large operators on the conversational agent market (B).
- Shaping market definition
9. The first part of the Consultation seeks to define the boundaries of the market for conversational agents. Providing the Authority with information regarding the market definition will help it assess competitive relations between market players and anticipate potential abuses from large vertically integrated players that could be detrimental to other players.
10. In this regard, the Authority asks whether conversational agents, chatbots and virtual assistants should be distinguished, who the main players are, and if and what barriers to entry or expansion are.
11. It makes particularly sense for operators that rely on external models to develop products or services, to respond to these questions. Many smaller players compete precisely because they can build on existing models[5] rather than developing their own.[6]
12. Explaining this dynamic helps the Authority understand how access conditions, interoperability, and dependency on upstream providers impact competition in the sector. Without such input, barriers to entry may be underestimated and potential risks such as self-preferencing by vertically integrated firms may not be fully captured.[7]
13. Closely linked to these access and dependency issues is the rapid multiplication of partnerships between developers of conversational agents and other technology or infrastructure providers.[8] The Consultation provides in this regard an opportunity for smaller players to ensure that the Authority understands both the benefits and the risks of these partnerships.
14. The Authority seeks to understand in what context stakeholders conclude partnerships and whether such partnerships may create competition concerns. These questions are not limited to major strategic partnerships, as they also cover commercial, technical, or distribution agreements that smaller players may rely on to access models, computing resources, or customers. Explaining the rationale behind such partnerships is therefore essential to ensure that the Authority distinguishes between arrangements that do not prevent market entry and those that may restrict competition.
15. With regard to both market definition and partnership assessment, responding to the Consultation will increase inter alia the likelihood that future policy and enforcement decisions will reflect the realities faced by smaller operators.
B. Clarifying competitive dynamics
16. Another set of questions asked by the Authority regards the integration of conversational agents into existing services offered by vertically integrated players. This can significantly reduce the visibility or viability of smaller operators. Participating in the Consultation allows smaller players, in particular, to explain whether such integrations create genuine competitive constraints or foreclose opportunities for third-party providers. Without input from smaller players, there is a risk that the Authority assesses these dynamics mainly through the lens of large platforms.
17. The Authority asks in this regard whether these forms integrations are likely to raise competition concerns and how interoperable vertically integrated services operate with third-party conversational agents. In practice, this means for the Authority better understanding whether and how independent providers can realistically connect to, compete with, or be distributed through dominant ecosystems, or whether technical, contractual, or commercial barriers limit those possibilities.
18. Market developments illustrate why stakeholders’ input is essential. The integration of generative AI into core products such as the deployment by Microsoft of AI assistants based on its own technology shows the bundling possibilities with existing dominant services.[9] These strategies may generate competitive constraints as shown in the recent decisions of the European Commission.[10]
19. For smaller market players, contributing to the Consultation serves several purposes. It allows them to provide their opinion as to whether third-party conversational agents can realistically interoperate with major platforms, to distinguish between legitimate product integration and potentially harmful tying practices, and to ensure that enforcement approaches are grounded in operational realities rather than assumptions.[11]
20. More broadly, the increasing integration of conversational agents into existing digital environments is also contributing to a deeper structural evolution of the sector, where these tools may progressively function less as mere conversational product and more as gateways to broader services. This dynamic leads directly to the next key issue examined by the Authority, the potential transformation of conversational agents into platforms.
21. Recent studies are showing that conversational agents are turning into fully fledged independent platforms.[12] Since this market is still emerging, new entrants have been able to compete with larger operators. However, to ensure that the market is competitive, providing information allowing the Authority to have a clear picture of competitive risks is vital.
22. Since conversational agents are becoming “meta-platforms”, they act as a new gateway for digital traffic. They can collect, filter and re-rank information before presenting it to consumers.[13]
23. On top of potential competitive risks between operators of conversational agents, online service providers might be affected by conversational agents turning into platforms. These new platforms could, in particular, restrict competition between online services through access restrictions, preferential treatment, or ecosystem lock-in.
24. While operators of conversational agents can help the Authority determine the relevant market for these new types of search engines, explaining potential barriers to entry and the practical impact on competition can also influence future enforcement and regulation by bringing their concerns about how these platforms function to the Authority’s attention.[14]
25. By clarifying how competition could be affected by the integration of conversational agents into existing services offered by vertically integrated companies and by conversational agents turning into meta-platforms, conversational agent operators and online retailers can ensure that their interest is fully taken into consideration by the Authority.
II. Emerging commercial uses of conversational agents
26. In addition to enquiring about the market structure, the Consultation addresses monetisation models and the use of conversational agents as advertisers (A). The Authority also seeks to get a better grasp of agentic commerce (B).
- Understanding better monetisation and advertising
27. The Authority noted in its previous report a lack of transparency regarding advertising and monetisation of AI.[15] The Authority is therefore asking market players and interested parties how monetisation is structured and evolving in the conversational agent market. It is also looking for more insight into advertising on conversational agents and how it could differ from other types of online advertising.
28. Participation of online service providers will be key in the conversational agent advertisement sector. As some academics have theorised, AI agents are creating a negative externality that could destroy the current ad-funded internet. Advertisers pay for human attention, when AI agents browse for consumers, they create “synthetic traffic” that has no purchasing intent.
29. Since advertisers cannot distinguish between human and AI traffic, they lower the price they pay for all ads. Publishers who rely on ad monetisation such as news or blog writers could lose a large part of their revenues. If too much traffic becomes AI-driven, publishers and other advertising space operators might get negatively impacted. Through their participation in the Consultation, advertisers could direct the attention of the Authority towards this issue and incentivise regulation.[16]
30. Furthermore, operators of conversational agents may want to answer to the Consultation’s questions about monetisation as they shape the Authority’s understanding of the revenue models in the sector. Non-vertically integrated agents rely on subscriptions, advertising, or other monetisation strategies to operate, and without input from market participants, the Authority may misunderstand these mechanisms or unintentionally restrict them. By contributing to the Consultation, operators will protect their business models, promote transparency, and influence regulation in a way that preserves fair competition and sustainable monetisation.
B. Protecting innovation on the agentic commerce market
31. Agentic commerce is a form of online commerce in which conversational agents make purchases “on behalf of” users.[17] As agentic commerce is a recent market, it is crucial that the Authority gets a realistic and clear picture of the competition risks it involves. Operators of such platforms as well as any online retailers affected have therefore an opportunity to voice their concerns and safeguard their interests in the context of the Consultation.
32. In this regard, the Authority is asking which sectors are the most likely to be impacted by agentic commerce, whether such platforms are likely to modify the demand of traditional e-commerce and if there are competitive risks on this market.
33. One of the main competitive challenges regarding agentic commerce is the potential two-sided algorithmic collusion. This is a situation where agentic commerce agents on both the supply side (sellers) and on the demand side (buyers) “learn” to cooperate to maintain high prices, effectively forming a digital cartel without any explicit human agreement.[18] Providing insight to the Authority into this type of commerce and its functioning will help it effectively approach such potential competitive disruption and ultimately limiting price inflations.
34. The agentic commerce market is rapidly evolving and large vertically integrated platforms are already in the process of cementing their power. For example, companies including Etsy, Shopify, and Walmart allow users to browse and purchase products directly through ChatGPT. A partnership with PayPal enables instant checkout through PayPal’s digital wallet, while PayPal processes merchant payments. [19]
35. Another area where stakeholders’ input would be particularly valuable regards the allocation of responsibility within agentic commerce ecosystems.
36. Conversational agents may generate inaccurate outputs or execute unintended transactions, uncertainty remains regarding the respective liability of operators, merchants and intermediaries. By explaining how these systems function in practice, stakeholders can help the Authority understand where risks arise and how responsibilities should realistically be distributed, thereby reducing legal uncertainty that could otherwise deter innovation.
37. Finally, participation in the Consultation can ensure that technical standards are not imposed too early creating costly compliance burdens or weaken business models. Instead, stakeholders can advocate for realistic solutions aligned with innovation incentives.
38. To conclude, the Consultation will directly influence how conversational agent markets are defined, how partnerships and integrations are assessed, and how emerging monetisation, advertising and agentic commerce models may be scrutinised in the future.
39. For smaller market players, technology integrators, retailers and service providers, in particular remaining silent carries the risk that future enforcement and regulatory approaches may be shaped primarily by the largest platforms, whose interests and market realities differ significantly from those of smaller players.
40. For all the stakeholders, responding to the Consultation allows them to explain their business models, highlight practical constraints, and ensure that the Authority (and potentially other competition authorities consequently) understand better how this market functions and develops. Persuasive submissions can reduce regulatory uncertainty, prevent disproportionate compliance burdens and protect access to key technologies or distribution channels.
[1] Autorité de la concurrence, The Autorité launches a public consultation on conversational agents, January 29, 2026, link.
[2] Autorité de la concurrence, Public consultation as part of the decision to start inquiries ex officio and issue an opinion on the conversational agents sector, January 29, 2026, link.
[3] Autorité de la concurrence, Avis 24-A-05, relatif au fonctionnement concurrentiel du secteur de l’intelligence articielle generative, June 28 2025, link.
[4] Autorité de la concurrence, The Autorité publishes its study on the competition issues surrounding the energy and environmental impact of artificial intelligence, December 17, 2025, link.
[5] Autorité de la concurrence, Generative artificial intelligence: the Autorité issues its opinion on the competitive functioning of the sector, June 28, 2024, link.
[6]Ibid, 5.
[7] Ibid, 7.
[8] L’usineDigitale, Le quotidien Le Monde s’associe à OpenAI, appâté par “des revenus significatifs”, March 14 2024 ,link and L’usineDigitale, Intelligence artificielle : l’Autorité de la concurrence s’autosaisit pour avis sur le secteur des agents conversationnels, January 12, 2026 ,link.
[9] Autorité de la concurrence, Generative artificial intelligence: the Autorité issues its opinion on the competitive functioning of the sector, June 28, 2024, link.
[10] European Commission, Press release, Commission notifies Meta of possible interim measures to reverse exclusion of third-party AI assistants from WhatsApp, February 9, 2025, link.
[11] Autorité de la concurrence, Avis 24-A-05, relatif au fonctionnement concurrentiel du secteur de l’intelligence artificielle générative, June 28 2025, link.
[12] Autorité de la concurrence, Conversational agents: the Autorité starts inquiries ex officio with a view to issuing an opinion, January 9, 2026 ,link and Mediametrie, Isabelle Lelouche Fillau, Young people at the forefront of conversational AI revolution, September 30, 2025, link.
[13] Shanghai Jiao Tong University, Jianghao Lin, Superplatforms Have to attack AI agents, May 23, 2025,link and Spears Chool of Business, Felipe M. Affonso, Vertical tacit collusion in AI-mediated markets, January 6, 2026, link
[14] Pub Affairs Bruxelles, Antitrust issues raised by answer engines, January 13,2026,link.
[15] The Platform law blog, Marc Barennes, Stefano Belle, Generative Artificial Intelligence and Competition: The Main Takeaways of the Recent Opinion Issued by the French Competition Authority, September 27, 2024 ,link
[16] Noth Dakota State University, Md Mahadi Hasan, AI Agents and the attention lemons problem in two-sided ad markets, July 31, 2025, link.
[17] Autorité de la concurrence, Public Consultation Questions, June 28, 2025, link.
[18] Singapore Management University, Anirban Mukherjee, Agentic AI: autonomy, accountability, and the algorithmic society, February 16, 2025, link. And Wilson Sonsini, European Antitrust Bimonthly bulletin, May/June 2025, link.
[19] Paypal, Press Release, OpenAI and PayPal Team Up to Power Instant Checkout and Agentic Commerce in ChatGPT (2025), October 28, 2025, link.
