On December 4, 2025, the European Commission formally opened an antitrust investigation into Meta Platforms’ rollout of artificial intelligence features in WhatsApp, escalating a growing clash between Brussels and Big Tech over who controls access to Europe’s fast‑growing AI markets. [1]
The probe zeroes in on “Meta’s new policy regarding AI providers’ access to WhatsApp”—rules that critics say could lock rival AI chatbots out of the world’s most popular messaging app while giving Meta’s own assistant, Meta AI, privileged access to more than two billion users worldwide. [2]
Key facts at a glance
- Who is investigating? The European Commission’s competition arm, led by Commissioner Teresa Ribera, acting under traditional EU antitrust rules (Article 102 TFEU), not the newer Digital Markets Act (DMA). [3]
- What’s under scrutiny? An October 2025 update to WhatsApp Business Solution terms that restricts third‑party AI providers from using WhatsApp when AI is their main service, while Meta’s own Meta AI remains integrated into the app. [4]
- Timeline:
- Parallel action: Italy’s antitrust watchdog AGCM is running its own case and has already moved toward interim measures targeting the same policy. [7]
- Potential penalty: Under EU competition law, Meta could face fines of up to 10% of its global annual turnover, which, based on its most recent revenue figures, could mean a theoretical maximum in the tens of billions of dollars. [8]
What triggered the EU’s Meta–WhatsApp AI investigation?
At the heart of the case is a policy change that, on its face, looks like a technical tweak to WhatsApp’s business terms—but regulators see it as potentially market‑shaping.
In October 2025, Meta updated the WhatsApp Business Solution Terms. Under the new rules, companies whose primary product is an AI service—for example, customer‑facing chatbots—are forbidden from using WhatsApp’s business communication tools. Businesses can still use some AI for ancillary functions like automated support, but not if the AI itself is the main value they offer. [9]
Critically, this doesn’t apply to Meta itself. While rival AI providers face exclusion, Meta’s own assistant, Meta AI, remains built into WhatsApp’s interface, accessible directly inside the app. [10]
According to the Commission and supporting analyses, the timeline looks like this: [11]
- 15 October 2025:
- New terms take effect for new AI providers; any company whose main product is an AI chatbot is blocked from onboarding to WhatsApp Business Solution.
- 15 January 2026:
- The same restriction kicks in for existing AI providers, giving current partners a grace period before being cut off.
The Commission’s concern is clear: by closing WhatsApp’s business APIs to most external AI providers while leaving its own assistant in place, Meta may be foreclosing access to a key digital “gatekeeper” and strengthening its dominant position in both messaging and AI assistant markets. [12]
Why Brussels says competition is at risk
The European Commission describes WhatsApp as a crucial communication channel for businesses and consumers across the EU, and notes that AI‑driven services embedded in messaging apps are becoming an important way to deliver customer support, personal assistants and shopping experiences. [13]
By tying AI access to WhatsApp so tightly to its own product, Meta is suspected of:
- Abusing a dominant position in communication services
- Italian regulators already argue that Meta enjoys a dominant position in app‑based communications and that the new terms could limit market access and technical development in AI chatbot services. [14]
- Foreclosing rivals from a critical distribution channel
- If AI providers cannot reach customers through WhatsApp, they lose access to tens of millions of European users who increasingly expect to interact with services through chat interfaces rather than websites or apps. [15]
- Creating a data advantage that competitors cannot replicate
- Meta AI, unlike rivals, can learn directly from interactions with WhatsApp’s vast user base. Italy’s antitrust authority has flagged that this exclusive training data could give Meta’s assistant a long‑term performance edge that persists even if rules are later changed. [16]
EU officials point out that AI markets in Europe are booming, with Italian regulators citing estimates of the bloc’s generative AI sector growing from about $4.4 billion in 2024 to $11.7 billion by 2026. That makes the design of platform access rules today especially consequential for future competition. [17]
Commissioner Teresa Ribera, who now oversees EU competition policy, has repeatedly signalled that she wants AI innovation to flourish without being crushed by a handful of dominant platforms. [18]
Italy’s early move put WhatsApp AI in the crosshairs
Long before Brussels acted, Italy’s competition authority (AGCM) had WhatsApp AI squarely in its sights.
- July 22, 2025:
AGCM opened a probe into Meta over allegations that integrating Meta AI into WhatsApp without proper consent could breach EU competition rules. [19] - October 15, 2025:
WhatsApp’s new Business Solution Terms were introduced, barring AI‑focused firms from using WhatsApp if AI is their main service. [20] - November 26, 2025:
The Italian authority broadened its investigation and launched interim proceedings, warning that the contractual changes and deeper integration of Meta AI could cause “serious and irreparable” harm to competition by locking rival chatbots out of WhatsApp while users become accustomed to Meta’s assistant. [21]
AGCM’s public documents highlight specific services that could be affected, including Microsoft Copilot, OpenAI’s ChatGPT, Perplexity and smaller European entrants that currently operate via WhatsApp. Once the grace period ends, many of these bots could effectively be forced off the platform in Italy. [22]
To avoid overlapping jurisdictions, the European Commission’s investigation covers the European Economic Area except Italy, which continues to handle the matter at national level. [23]
Why this case matters for Europe’s wider tech crackdown
The WhatsApp AI case arrives amid an aggressive EU push to rewrite the rules for Big Tech:
- The Digital Markets Act (DMA) labels certain platforms as “gatekeepers” and imposes strict obligations on how they treat business users and competitors.
- In April 2025, Meta was fined €200 million under the DMA for its controversial “consent or pay” model, which forced users to choose between targeted ads and a paid, tracking‑free subscription. [24]
- Other tech giants, including Apple, Amazon and Microsoft, are also facing DMA enforcement and traditional antitrust probes over app distribution, cloud services and bundling practices. [25]
Notably, the Commission chose to pursue the WhatsApp policy under classic antitrust rules, not the DMA. [26]
That choice matters:
- Antitrust law (Article 102 TFEU) allows the EU to investigate abuse of dominance on a case‑by‑case basis, potentially leading to heavy fines and behavioural remedies tailored to the specific conduct. [27]
- The Commission may be signalling that gatekeeper rules alone aren’t enough when a platform allegedly uses its dominance to carve out special privileges for its own AI while closing the door on others.
In short, WhatsApp AI is emerging as a test case for how Europe will respond when powerful platforms use control over critical interfaces—like messaging apps—to shape which AI tools users can access.
Meta’s response: “Baseless claims” and system strain
Meta insists it has done nothing wrong.
A WhatsApp spokesperson has said the antitrust allegations are “baseless”, arguing that a sudden influx of external chatbots puts technical strain on systems “not designed to support” such usage. The company also stresses that the AI space remains intensely competitive, with people able to access numerous AI services through app stores, search engines, email, other messaging platforms and operating systems. [28]
In broad terms, Meta’s defence rests on three points:
- Technical justification
- The company says WhatsApp’s infrastructure wasn’t built to host large numbers of third‑party AI agents at scale and that limiting such integrations protects reliability and safety.
- Plenty of alternatives
- Meta maintains that users and businesses can still reach rival AI services through other channels—web, native apps, or integrations in different platforms—so the policy does not meaningfully restrict competition. [29]
- Highly competitive AI market
- With tech giants and start‑ups worldwide racing to launch chatbots and enterprise assistants, Meta argues that no single service, including Meta AI, can dominate the field simply by being integrated into WhatsApp.
Regulators, however, are likely to scrutinize how realistic those alternatives are in practice for European SMEs that rely on WhatsApp to reach customers, and whether Meta’s own assistant enjoys advantages others cannot match.
What penalties Meta could face
Under EU competition law, the European Commission can impose fines of up to 10% of a company’s global turnoverfor violations of Articles 101 or 102 TFEU. [30]
Meta’s financial scale makes that ceiling substantial:
- Meta’s revenue for 2023 was about $134.9 billion. [31]
- Public financial data suggest revenue rose to roughly $164.5 billion in 2024, with trailing‑twelve‑month revenue around $189.5 billion in 2025. [32]
On those numbers, a theoretical maximum fine could approach $16–19 billion, though fines usually come in lower and depend on the gravity and duration of the infringement, as well as mitigating factors such as cooperation. [33]
Beyond monetary sanctions, the Commission could order behavioural remedies, such as:
- Re‑opening WhatsApp Business Solution to AI providers on fair, transparent and non‑discriminatory terms.
- Requiring clearer separation between Meta AI and the platform’s core messaging functions.
- Imposing safeguards around data access and use, so Meta cannot leverage WhatsApp interaction data in ways denied to competing assistants.
In extreme cases, EU law allows for structural remedies, but these are rare and would likely only be considered if behavioural measures were deemed insufficient.
What’s at stake for AI start‑ups, enterprises and users
For AI developers, particularly in Europe, the WhatsApp case is about more than one platform’s terms of service. It goes to the heart of who controls digital distribution in the AI era.
- AI start‑ups and SaaS providers
- Many emerging AI companies rely on WhatsApp bots to reach users who prefer to interact in chat rather than through standalone apps. Losing that channel could significantly raise customer acquisition costs and entrench incumbents that already enjoy scale. [34]
- Enterprises and SMEs
- European businesses increasingly use AI agents on WhatsApp for customer support, lead qualification and basic sales. If only Meta’s assistant can operate natively inside WhatsApp, firms may have less choice and face higher switching costs than if multiple providers were allowed to compete on equal terms.
- Consumers
- For end users, the risk is reduced diversity of AI services. If Meta AI becomes the default (and sometimes only) assistant inside a dominant messaging platform, people may be less likely to seek out alternatives—even if those alternatives offer stronger privacy protections, language support or niche features.
Italy’s AGCM has warned of “lock‑in effects”, where users’ growing familiarity with Meta AI, reinforced by personalization and data advantages, could make it difficult for rivals to catch up later—even if regulators eventually intervene. [35]
The case therefore touches on innovation policy as much as competition law: whether Europe can cultivate a vibrant AI ecosystem if one gateway platform tightly controls who is allowed to build on top of it.
What happens next
The Commission has said it will treat the WhatsApp AI case as a priority, but, as is common with EU antitrust probes, there is no formal deadline for completion. [36]
Typical steps include:
- Information‑gathering
- Requests for documents and data from Meta and third‑party AI providers.
- Market testing
- Surveys and consultations with businesses and regulators across the EU, including coordination with national authorities like Italy’s AGCM. [37]
- Statement of objections
- If the Commission finds indications of abuse, it can issue a formal statement of objections laying out its theory of harm, to which Meta can respond.
- Potential interim measures
- Given that the strictest aspects of Meta’s policy are due to hit existing AI providers from 15 January 2026, regulators may consider temporary measures to prevent irreversible market effects while the case is ongoing—mirroring the interim proceedings already opened in Italy. [38]
Whatever the precise timeline, the investigation ensures that Meta’s WhatsApp AI strategy will remain under intense regulatory and industry scrutiny well into 2026 and beyond.
Why this case could set a global precedent
For Meta, the WhatsApp AI investigation is unfolding just weeks after internal documents revealed the company is already facing mounting regulatory pressure over its broader use of AI and its role in online fraud, and months after it was fined under the DMA. [39]
For regulators worldwide, the case may become a template for tackling AI‑related gatekeeping:
- If the EU forces WhatsApp to open up to rival AI assistants, other jurisdictions could adopt similar approaches to ensure competition on messaging platforms.
- If Meta successfully defends the policy, Big Tech companies may feel emboldened to bake proprietary AI more deeply into core services while limiting third‑party access.
In that sense, Brussels isn’t just deciding the future of AI inside WhatsApp. It’s testing how far regulators are willing—and able—to go when the interface where people talk to each other becomes the place where they talk to AI.
References
1. www.reuters.com, 2. www.reuters.com, 3. www.reuters.com, 4. www.investing.com, 5. www.reuters.com, 6. www.investing.com, 7. ppc.land, 8. eur-lex.europa.eu, 9. www.investing.com, 10. www.reuters.com, 11. www.investing.com, 12. www.reuters.com, 13. www.reuters.com, 14. ppc.land, 15. ppc.land, 16. ppc.land, 17. ppc.land, 18. en.wikipedia.org, 19. www.dlapiper.com, 20. ppc.land, 21. ppc.land, 22. ppc.land, 23. www.investing.com, 24. en.wikipedia.org, 25. www.bclplaw.com, 26. www.reuters.com, 27. eur-lex.europa.eu, 28. www.reuters.com, 29. www.reuters.com, 30. eur-lex.europa.eu, 31. www.macrotrends.net, 32. companiesmarketcap.com, 33. eur-lex.europa.eu, 34. ppc.land, 35. ppc.land, 36. www.investing.com, 37. www.dlapiper.com, 38. ppc.land, 39. www.reuters.com
