Italy’s antitrust authority has escalated its investigation into Meta Platforms, expanding a probe that began in July 2025 to now examine how the tech giant’s AI integration and new business policies on WhatsApp could be stifling competition in the rapidly growing artificial intelligence market. The Autorità Garante della Concorrenza e del Mercato (AGCM) announced on November 26, 2025, that it is broadening its investigation while simultaneously initiating procedures for potential interim measures that could freeze Meta’s recent policy changes.
This development represents a critical juncture in the intersection of AI innovation, platform dominance, and regulatory oversight. With WhatsApp serving more than 37 million users in Italy alone and Meta’s global turnover reaching $164.5 billion in 2024, the stakes are extraordinarily high. If found guilty of abusing its dominant position under EU competition rules, Meta could face fines of up to 10% of its worldwide revenue, potentially totaling more than $16 billion.
Understanding the Investigation: From July Origins to November Expansion
The Initial Probe: July 2025
The AGCM launched its initial investigation into Meta on July 30, 2025, following inspections at the company’s Italian offices conducted with assistance from Italy’s tax police special antitrust unit. The original probe centered on allegations that Meta had violated Article 102 of the Treaty on the Functioning of the European Union (TFEU) by integrating its Meta AI assistant into WhatsApp without obtaining proper user consent.
Article 102 TFEU prohibits companies from abusing a dominant market position. Under this framework, holding a dominant position is not illegal in itself. However, companies with market dominance carry a special responsibility not to engage in conduct that impairs genuine competition. The provision allows the European Commission and national competition authorities to investigate and penalize anticompetitive behavior, with fines reaching up to 10% of a company’s global annual turnover.
Meta AI was integrated into WhatsApp in March 2025, appearing prominently in the app’s search bar and interface. The AI assistant, powered by Meta’s Llama 4 architecture, offers chatbot-style responses, can generate images on demand, answer questions, draft messages, and provide various virtual assistant functions. Users access it through the dedicated Meta AI chat feature or by using “@Meta AI” mentions within conversations.
The AGCM’s concern centered on what antitrust law calls “tying,” a practice where a company leverages a popular product to force adoption of another service. By automatically integrating Meta AI into WhatsApp without explicit user consent, regulators argued that Meta was using its dominant position in messaging to gain an unfair advantage in the emerging AI assistant market.
The November Broadening: New Business Terms Under Scrutiny
On November 26, 2025, the AGCM announced a significant expansion of its investigation. The probe now encompasses two additional concerns beyond the original Meta AI integration issue.
First, the authority is examining Meta’s October 15, 2025, changes to the WhatsApp Business Solution Terms. These updated terms introduced a new clause explicitly prohibiting “AI providers” from using the WhatsApp Business API when AI services constitute the primary functionality being offered. This policy effectively bans developers of large language models, generative AI platforms, and general-purpose AI assistants from distributing their services through WhatsApp’s business platform.
The new restriction applies immediately to companies attempting to join WhatsApp’s business platform and takes effect on January 15, 2026, for existing providers. Major AI services including OpenAI’s ChatGPT, Microsoft’s Copilot, and Perplexity have already announced they will discontinue WhatsApp integration by the January deadline.
Second, the investigation covers Meta’s continued integration of new Meta AI interaction tools and features into WhatsApp. The regulator notes that while restricting third-party AI access, Meta has simultaneously been expanding the visibility and functionality of its own AI assistant within the messaging platform.
The Competitive Landscape: WhatsApp’s Dominant Position in Italy
Understanding the investigation’s significance requires examining WhatsApp’s market position. In Italy, WhatsApp commands a penetration rate of approximately 97% among messaging app users, making it one of the most dominant messaging platforms in the country. With more than 37 million Italian users, WhatsApp has become an essential communication infrastructure that people rely on daily for personal and business interactions.
Globally, WhatsApp reached approximately 3 billion monthly active users as of 2025, making it the world’s most popular messaging application. The platform is available in over 180 countries and supports 60 languages. In 2024, WhatsApp users exchanged an estimated 140 billion messages daily, representing approximately 97 million messages every minute.
This massive user base creates what regulators call “network effects,” where the platform’s value increases as more people use it. These effects create high switching costs for users. Changing to a different messaging service means potentially losing contact with friends, family, and business associates who remain on WhatsApp. The AGCM specifically highlighted consumer reluctance to change communication habits as a factor that could entrench Meta’s competitive advantages.
Meta AI Integration: Technical Features and Competitive Implications
How Meta AI Works Within WhatsApp
Meta AI operates on Llama 4, Meta’s latest large language model that features a mixture-of-experts design for faster inference, long-context windows capable of remembering up to 200,000 tokens per session, vision-aware variants for analyzing shared photos, and multilingual fluency optimized for over 100 languages.
The integration appears in multiple ways within WhatsApp. Users can access a dedicated Meta AI chat thread for private conversations with the assistant. They can use @Meta AI mentions inside group chats to ask questions visible only to tagged participants. The system provides smart reply surfaces with inline suggestions when drafting messages, similar to Gmail’s Smart Compose feature.
One standout capability is instant image generation. Users can type prompts starting with “imagine” to create AI-generated images within seconds, all without leaving the WhatsApp conversation. The assistant can also answer factual questions, help with everyday planning, write creative content like poems or messages, and provide information on various topics.
In June 2025, Meta introduced message summaries using what it calls “Private Processing” technology. This optional feature uses Meta AI to summarize unread messages in chats, with the summary visible only to the individual user. Meta emphasizes that this technology allows AI processing without Meta or WhatsApp seeing actual message contents, maintaining end-to-end encryption for non-AI interactions.
The Competitive Advantage Concern
The AGCM’s investigation focuses on whether this deep integration gives Meta an insurmountable competitive advantage. By embedding Meta AI directly into an application used by 37 million Italians, Meta effectively places its AI assistant in front of a captive audience without users needing to download separate applications or create new accounts.
This integration creates several competitive barriers. First, it establishes Meta AI as the default option for users seeking AI assistance within their familiar messaging environment. Second, it allows Meta to collect valuable interaction data that can improve its AI models. Third, it creates switching costs where users who become familiar with Meta AI’s interface and capabilities may be reluctant to try competing services.
Regulators warn of a potential “lock-in effect” where users become functionally dependent on Meta AI. As the assistant learns from interactions and provides increasingly personalized and useful responses, the cost of switching to competitors rises. This effect is particularly pronounced because competing AI services now face barriers to distributing through WhatsApp, the platform where users already spend significant time.
According to user data, the average WhatsApp user spends approximately 38 minutes per day on the application, totaling about 19.4 hours per month. This sustained engagement makes WhatsApp an incredibly valuable distribution channel for any service, including AI assistants.
The WhatsApp Business Solution Terms: Blocking Third-Party AI
What Changed on October 15, 2025
Meta’s updated WhatsApp Business Solution Terms introduced a specific prohibition targeting what it defines as “AI Providers.” The new clause states that providers and developers of artificial intelligence or machine learning technologies, including large language models, generative AI platforms, and general-purpose artificial intelligence assistants, are strictly prohibited from accessing or using the WhatsApp Business Solution when such technologies constitute the primary (rather than incidental or ancillary) functionality being offered.
This language is deliberately broad. It gives Meta discretion to determine what qualifies as “primary functionality” versus “incidental or ancillary” use. The policy allows businesses to continue using AI for customer service automation, order confirmations, support triage, and transactional workflows. However, it explicitly blocks general-purpose conversational AI assistants from using WhatsApp as a distribution channel.
The enforcement timeline creates urgency. New AI providers are immediately prohibited from joining the platform. Existing providers, including major services like ChatGPT, Copilot, and Perplexity, must cease operations by January 15, 2026. This narrow migration window gives affected companies approximately two months to transition users to alternative platforms.
Meta’s Justification
WhatsApp has defended the policy change on technical and operational grounds. A company spokesperson stated that the WhatsApp API business interface “was never designed to be used for AI chatbots and doing so would place severe strain on our systems.”
Meta argues that general-purpose AI chatbots generate unpredictable, high-volume message patterns that differ substantially from the structured, predictable messages the Business API was designed to handle. Open-ended conversational AI can produce long session histories, varied content requiring intensive moderation, and infrastructure demands that exceed the platform’s intended use case.
The company maintains that the restriction preserves the Business Solution for its intended purpose: helping companies manage customer communications, send notifications, provide support, and conduct business transactions. Meta emphasizes that the change “does not affect the tens of thousands of businesses who provide customer support and send relevant updates, or the businesses using the AI assistant of their choice to chat with their customers.”
From Meta’s perspective, the policy protects system integrity, maintains service quality for legitimate business users, and prevents resource strain from applications that were never part of the platform’s design specifications.
Regulatory Concerns and Competitive Effects
The AGCM sees the situation differently. In its November announcement, the authority stated that the contractual changes “could limit production, market access or technical developments in the AI Chatbot services market, to the detriment of consumers, and may amount to a possible violation of Article 102 TFEU.”
Regulators argue that WhatsApp’s massive Italian user base, combined with consumer reluctance to change communication habits, creates a situation where blocking third-party AI access effectively shuts competitors out of a crucial market. The timing raises additional concerns, as Meta implements these restrictions while simultaneously expanding its own Meta AI integration.
This creates what antitrust experts call a “self-preferencing” scenario. A vertically integrated dominant company (Meta owns both the WhatsApp platform and the Meta AI service) uses control over essential infrastructure (WhatsApp’s messaging platform and Business API) to favor its own offerings (Meta AI) while excluding rivals (ChatGPT, Copilot, and others).
The investigation must determine whether Meta’s technical justifications are legitimate or whether they serve as a pretext for anticompetitive exclusion. The AGCM will likely examine whether less restrictive alternatives exist, such as implementing usage limits, charging higher fees for AI traffic, or establishing verified partner programs that could accommodate both system integrity and competitive access.
The Interim Measures Procedure: Urgent Regulatory Action
Understanding Section 14-bis of Law 287/1990
The AGCM’s announcement that it has initiated a procedure for potential interim measures represents a significant escalation. Under Section 14-bis of Italian Law 287/1990, the competition authority can impose emergency injunctions to prevent serious and irreparable harm to competition while a full investigation continues.
Interim measures are reserved for situations where waiting for the conclusion of a standard antitrust probe would be insufficient because the market structure could be permanently altered. The authority must demonstrate three elements: urgency, a risk of serious and irreparable harm, and a prima facie case of infringement under competition law.
If granted, interim measures could include suspending the new WhatsApp Business Solution Terms, preventing Meta from further integrating Meta AI features into WhatsApp, requiring Meta to provide temporary access to third-party AI providers, or mandating specific changes to how Meta AI appears within the WhatsApp interface.
Why Interim Measures May Be Warranted
The AGCM’s decision to pursue interim measures suggests the authority believes waiting until the investigation’s scheduled conclusion on December 31, 2026, could allow irreversible competitive harm.
Once third-party AI assistants are removed from WhatsApp on January 15, 2026, users who relied on these services will need alternatives. Many will likely migrate to Meta AI simply because it remains conveniently accessible within their existing WhatsApp workflow. This user migration could establish Meta AI’s market position so firmly that even if the final investigation finds violations, restoring competition becomes extremely difficult.
Network effects and data advantages compound this concern. As Meta AI accumulates more users and interactions, its language models improve through training on this data. Competing services denied access to WhatsApp users miss this improvement cycle. The gap between Meta AI’s capabilities and those of competitors could widen to the point where regulatory remedies cannot restore meaningful competition.
The AGCM noted that Meta’s conduct is “capable of severely and irreparably undermining the contestability of the market, due to consumers’ limited propensity to change their habits, which hampers switching to competing services.”
Potential Outcomes
If the authority imposes interim measures, Meta would face immediate operational impacts. The company might be required to extend the January 15, 2026, deadline, allowing third-party AI services to continue operating on WhatsApp during the investigation. It could be forced to modify or clarify the “AI Providers” restriction to create pathways for verified AI services that meet specific technical requirements.
Alternatively, the AGCM might require Meta to offer users explicit opt-in or opt-out choices for Meta AI integration, ensuring that the assistant’s presence does not constitute forced bundling. The authority could mandate that Meta AI appear less prominently in WhatsApp’s interface or include notifications about alternative AI services.
Meta will have opportunities to contest any interim measures. The company can argue that its technical justifications are legitimate, that alternative distribution channels exist for AI services (mobile apps, websites, other messaging platforms), and that the measures would impose disproportionate costs or technical challenges.
Broader EU Competition Law Context
Article 102 TFEU and Abuse of Dominance
Article 102 TFEU forms the legal foundation for this investigation. The provision states: “Any abuse by one or more undertakings of a dominant position within the internal market or in a substantial part of it shall be prohibited as incompatible with the internal market in so far as it may affect trade between Member States.”
Examples of prohibited abuse include imposing unfair purchase or selling prices, limiting production or technical development to consumer prejudice, applying different conditions to equivalent transactions, and making contracts subject to acceptance of supplementary obligations unrelated to the contract’s subject.
The European Commission’s guidance on Article 102 enforcement emphasizes that dominant companies have a special responsibility not to impair genuine competition. While dominance itself is not illegal, dominant firms face restrictions that non-dominant competitors do not.
To establish a violation, authorities must prove three elements. First, the relevant market must be defined (here, both the messaging market and the AI assistant market). Second, the company must hold a dominant position in that market (WhatsApp’s 97% penetration rate in Italy clearly establishes dominance in messaging). Third, the company must have engaged in abusive conduct (the question at the heart of this investigation).
Parallels to Other Meta Competition Cases
This investigation follows other significant competition enforcement actions against Meta in Europe. In November 2024, the European Commission fined Meta approximately 800 million euros for integrating Facebook Marketplace into Facebook’s social network in a way that gave Meta an unfair advantage over rival classified ad services. Meta is currently appealing that decision.
The Facebook Marketplace case centered on similar tying allegations. By automatically including Marketplace in Facebook’s interface and linking Marketplace listings to Facebook user profiles, Meta allegedly leveraged its social network dominance to foreclose competition in online classified ads.
The FTC in the United States is pursuing litigation seeking to break up Meta based on concerns that it excluded competitors through its 2012 acquisition of Instagram and 2014 acquisition of WhatsApp. Mark Zuckerberg testified in this case in April 2025. While the FTC case focuses on past acquisitions, it reflects ongoing regulatory concerns about Meta’s approach to competition.
In March 2024, the European Commission opened formal proceedings regarding Meta’s “subscription for no ads” model, questioning whether it complies with the Digital Markets Act. The Commission issued preliminary findings in July 2024 suggesting the model does not meet regulatory requirements, with a final decision expected in March 2025.
These cases collectively illustrate that Meta operates under intense competitive scrutiny across multiple jurisdictions and product areas. The Italian WhatsApp AI investigation fits within a broader pattern of regulatory challenges to Meta’s business practices.
The Emerging AI Competition Landscape
The Race for AI Distribution Channels
The WhatsApp investigation occurs amid an intense competitive race in artificial intelligence. Major technology companies are investing billions in AI development, creating increasingly sophisticated large language models and generative AI systems.
However, technical capabilities alone do not guarantee market success. Distribution matters enormously. Even the most advanced AI assistant has limited value if users cannot easily access it. This reality makes established platforms with massive user bases, like WhatsApp, extremely valuable distribution channels.
Meta’s global family of apps reported 3.35 billion daily active people as of December 2024, with WhatsApp being a core component of this ecosystem. In 2024, Meta generated $164.5 billion in revenue, with 99% coming from advertising across its platforms. The company has made AI integration a central strategic priority, with CEO Mark Zuckerberg stating that Meta will continue investing significantly in AI infrastructure.
For competing AI companies, access to messaging platforms provides direct, frictionless user engagement. Users can interact with AI without downloading separate apps, creating accounts, or changing their daily habits. This convenience translates to higher adoption rates and more frequent usage.
Microsoft’s Copilot, for example, used WhatsApp integration to reach millions of users who might not otherwise download Copilot’s standalone app. OpenAI similarly benefited from WhatsApp distribution for ChatGPT. Losing this channel forces these companies to rely on their own apps, websites, and other platforms where they must compete for user attention against countless alternatives.
Data Advantages and Model Improvement
Beyond distribution, interaction data from WhatsApp could provide Meta with significant AI development advantages. Large language models improve through training on vast amounts of text data. Conversational data, showing how real users phrase questions and what information they seek, is particularly valuable for refining AI responses.
While Meta emphasizes that WhatsApp messages remain end-to-end encrypted and that Meta AI processing occurs separately, the company acknowledges that interactions with Meta AI are not end-to-end encrypted and are visible to Meta for service improvement. This creates a data asymmetry. Meta can train its models on WhatsApp-based Meta AI interactions, while competitors no longer have access to this user engagement channel.
The AGCM noted this concern in its investigation documents, stating that integrating Meta AI into WhatsApp facilitates “rapid user base expansion and potentially training AI models on user interactions, thereby generating exclusionary effects.”
As AI capabilities increasingly depend on training data quality and quantity, this data advantage could translate into technical superiority. If Meta AI becomes demonstrably better than competitors due to superior training data, the competitive gap widens beyond what regulatory intervention can easily remedy.
Impact on Major AI Providers
Microsoft Copilot’s WhatsApp Departure
Microsoft introduced Copilot on WhatsApp in late 2024 as a low-friction way for people to interact with its AI assistant within a familiar messaging environment. Users could message Copilot like any other contact to get summaries, ideas, quick drafts, and answers without installing separate applications or signing into Microsoft accounts.
According to Microsoft, this WhatsApp deployment reached millions of users before Meta’s policy change. The company announced that Copilot will discontinue WhatsApp support on January 15, 2026, directing users to migrate to Copilot’s mobile apps (iOS and Android), the web interface at copilot.microsoft.com, or the integrated Copilot experiences within Windows.
Microsoft framed this as a platform policy enforcement rather than a Microsoft product choice, emphasizing that the decision reflects Meta’s contractual restrictions. However, losing WhatsApp distribution represents a significant setback. Microsoft must convince users to change their behavior, downloading new apps or visiting websites instead of continuing seamless WhatsApp-based interactions.
OpenAI’s ChatGPT and Other Services
OpenAI updated its help documentation to inform users that ChatGPT will stop working on WhatsApp after January 15, 2026. Like Microsoft, OpenAI benefited from WhatsApp’s distribution reach, allowing users to access ChatGPT’s capabilities through a familiar messaging interface.
Perplexity, an AI-powered answer engine, also used WhatsApp integration to reach users. Smaller AI startups and emerging services face even greater challenges, as WhatsApp provided an accessible distribution channel that leveled the competitive playing field somewhat against tech giants with established user bases.
For all these providers, the January 2026 deadline creates urgency. They must migrate users, establish alternative engagement channels, and potentially adjust business models built around multi-platform accessibility. The migration period coincides with what many analysts expect to be a critical phase in AI market development, when user habits and preferences solidify.
Market Dynamics and Consumer Impact
What Italian WhatsApp Users Experience
For the average Italian WhatsApp user, Meta AI’s integration represents both convenience and potential limitation. The convenience stems from having AI assistance available without switching apps or creating new accounts. Users can ask Meta AI questions, generate images, get writing help, or receive information without interrupting their messaging workflow.
However, the limitation emerges from reduced choice. Previously, users could access multiple AI assistants through WhatsApp by messaging different services. ChatGPT, Copilot, and Perplexity each offered distinct capabilities, interfaces, and strengths. Power users could select the AI best suited for specific tasks. ChatGPT might excel at creative writing, Copilot at productivity tasks, and Perplexity at research-style queries.
After January 15, 2026, this choice disappears within WhatsApp. Meta AI becomes the sole general-purpose AI assistant accessible through the platform. Users seeking alternative services must leave WhatsApp, open separate apps or websites, and potentially duplicate information they already shared in WhatsApp conversations.
The AGCM’s investigation centers on whether this reduction in choice harms consumers. Does the convenience of integrated Meta AI outweigh the loss of competitive options? Would users prefer to choose from multiple AI assistants within WhatsApp? Would competition between multiple in-app AI services drive innovation and quality improvements that Meta AI alone might not achieve?
Business Users and the WhatsApp Business API
WhatsApp Business Solution serves a different user segment: companies using WhatsApp for customer communications, support, notifications, and commerce. According to Meta, over 200 million businesses use WhatsApp Business monthly as of July 2024.
These business users are less affected by the AI provider restriction, according to Meta’s statements. Companies can continue using AI for customer service automation, chatbots that answer frequently asked questions, automated booking systems, order status updates, and similar business functions.
The distinction Meta draws is that these uses constitute “incidental or ancillary” AI functionality supporting core business operations, not the “primary functionality” that the prohibition targets. A hotel using AI to handle reservation queries remains acceptable. An AI assistant company using WhatsApp to deliver its service does not.
However, this distinction may create confusion and compliance challenges. Where exactly is the line between permitted business automation and prohibited general-purpose AI? If a company develops an AI assistant that both handles customer service and provides general information, which category applies? Meta retains discretion to make these determinations, which could lead to inconsistent enforcement or self-censorship by businesses uncertain about compliance.
Meta’s Strategic Position and Response
The Financial Stakes
With annual revenue of $164.5 billion in 2024 and the potential for fines reaching 10% of worldwide turnover, Meta faces maximum penalties exceeding $16 billion if found guilty of Article 102 violations. This potential liability represents one of the largest possible antitrust fines in technology industry history.
Beyond direct fines, adverse findings could force Meta to restructure its WhatsApp AI integration, unbundle Meta AI from WhatsApp, provide equal access to third-party AI services, or implement other remedies that fundamentally alter its competitive strategy.
Meta’s stock market valuation and investor expectations reflect the company’s AI ambitions. The company has positioned AI integration across its platforms as a core growth strategy. Restrictions on this integration could impact investor confidence and future revenue projections.
Meta’s Defense Strategy
Meta has emphasized cooperation with Italian authorities while defending its conduct on multiple grounds. The company argues that offering free AI access through WhatsApp benefits users by making advanced technology widely available through familiar, trusted platforms.
On the Business Solution terms, Meta stresses technical necessity. The WhatsApp infrastructure was designed for specific use cases, and general-purpose AI chatbots impose burdens that could affect platform stability and service quality for all users. Preventing system strain is a legitimate business justification, not anticompetitive exclusion.
Meta also contends that alternative distribution channels exist for AI services. Competing AI providers can offer mobile apps, websites, integration with other messaging platforms, and direct distribution strategies. WhatsApp is one channel among many, not an essential facility that competitors cannot do without.
Furthermore, Meta will likely argue that its Meta AI integration represents product innovation and investment that benefits consumers. Companies should be free to improve their products with new features. Requiring platform owners to provide equal access to every potential competitor could discourage innovation and investment.
Regulatory Implications Beyond Italy
Potential for EU-Wide Action
While the AGCM investigation is formally an Italian national proceeding, it carries EU-wide implications. The authority stated it is acting “in close cooperation with the relevant offices of the European Commission,” suggesting coordination at the European level.
Other EU national competition authorities may launch parallel investigations. The European Commission itself could open a formal proceeding under Article 102 TFEU, particularly if the Italian investigation produces findings of widespread harm to the internal market.
The Digital Markets Act (DMA), which came into force in 2023, designates certain large platforms as “gatekeepers” subject to specific obligations. These include prohibitions on self-preferencing and requirements to allow interoperability. While the DMA focuses on ex-ante regulation (preventing problems before they occur) rather than ex-post enforcement (punishing violations after the fact), it provides additional regulatory tools that could apply to Meta’s conduct.
If Meta’s WhatsApp practices are found to violate Article 102, the Commission could pursue both antitrust fines and DMA enforcement, potentially creating cumulative penalties and remedial obligations.
Global Regulatory Attention
Competition authorities outside Europe will watch this case closely. The United Kingdom’s Competition and Markets Authority (CMA) has shown increased focus on AI market dynamics and digital platform dominance. The U.S. Federal Trade Commission and Department of Justice have both initiated major technology antitrust cases, though U.S. enforcement standards differ from EU approaches.
Regulatory precedents from the Italian investigation could influence how authorities worldwide approach similar issues: platform owners integrating their own services while restricting competitors, the role of technical justifications in exclusionary conduct analysis, the application of traditional competition law to emerging AI markets, and the appropriate balance between innovation incentives and competitive openness.
Setting Precedents for AI Platform Regulation
The WhatsApp investigation represents one of the first major competition law cases examining how dominant platforms should handle AI integration. The outcome will establish important precedents for questions that will recur across the technology industry.
Should platform owners be allowed to integrate AI capabilities while excluding third-party AI services? What technical justifications are sufficient to justify different treatment? How do regulators balance innovation encouragement with competitive fairness? When do data advantages from platform integration become anticompetitive barriers?
The answers to these questions will shape not only Meta’s conduct but also how Google, Apple, Microsoft, Amazon, and other platform operators approach AI integration in their ecosystems. A finding against Meta could lead to industry-wide requirements for competitive AI access, while a decision favoring Meta could validate integrated approaches.
Timeline and Next Steps
Investigation Milestones
The AGCM investigation follows a structured timeline governed by EU and Italian competition law. The proceeding is scheduled to conclude by December 31, 2026, giving the authority approximately 13 months from the November 2025 expansion to complete its analysis and issue findings.
During this period, Meta and other interested parties have 60 days from official notification to exercise their right to be heard, submit written defenses, and present evidence. The authority will gather additional evidence, potentially including document requests, interviews, market studies, and economic analysis.
Interim measures, if imposed, could take effect much sooner. The AGCM stated that Meta must present defensive briefs within only seven days of notification regarding the interim measures procedure, reflecting the urgency the authority assigns to preventing immediate competitive harm.
January 15, 2026: The Critical Date
January 15, 2026, represents a pivotal moment. This is when Meta’s WhatsApp Business Solution Terms requiring third-party AI services to cease operations take effect. Major providers including Microsoft Copilot and OpenAI’s ChatGPT have confirmed they will discontinue WhatsApp integration on this date.
If the AGCM imposes interim measures before this deadline, the January 15 enforcement could be suspended pending the investigation’s outcome. Third-party AI services might continue operating on WhatsApp during the probe, preserving the competitive status quo.
If no interim measures are imposed, January 15, 2026, will mark a significant market structure change. Third-party AI services will exit WhatsApp, potentially triggering user migrations to Meta AI and establishing market patterns that could prove difficult to reverse even if the final investigation finds violations.
Possible Outcomes
Several scenarios could emerge from this investigation. The AGCM could find no violation, determining that Meta’s conduct has legitimate business justifications and does not constitute abuse of dominance. This outcome would allow Meta to maintain its current policies, though it seems less likely given the authority’s decision to expand the investigation and pursue interim measures.
The authority could find violations and impose fines while requiring specific remedies. Potential remedies might include requiring Meta to allow third-party AI services access to WhatsApp Business Solution under reasonable terms, mandating that Meta AI integration must be optional with clear opt-out mechanisms, requiring equivalent positioning for third-party AI services if Meta continues to promote Meta AI within WhatsApp, or ordering structural separation between WhatsApp’s messaging infrastructure and Meta’s AI services.
A negotiated settlement represents another possibility. Meta could agree to modify its practices voluntarily in exchange for the authority closing the investigation without formal violation findings. Settlements often include commitments monitored over several years to ensure compliance.
Implications for the Future of AI Competition
The Battle for AI Distribution
This investigation exemplifies a fundamental tension in modern technology markets. Companies that built dominant platforms in one era (social networking, messaging, search) now seek to leverage those platforms for competitive advantage in the next era (artificial intelligence).
Platform owners argue they should be free to innovate, integrating new technologies into their services to benefit users. Regulators worry that this integration allows incumbents to foreclose emerging competitors, preventing the dynamic competition that drives innovation and consumer benefits.
The stakes extend beyond one investigation or one company. How regulators resolve these tensions will shape which companies succeed in AI markets, whether new entrants can compete against established platforms, and ultimately whether AI develops as a diverse competitive ecosystem or consolidates under a few dominant players.
User Agency and Platform Power
At a deeper level, the investigation raises questions about user agency in platform ecosystems. When users choose WhatsApp for messaging, are they implicitly choosing Meta for all related services? Should users expect to access multiple providers’ services through a single platform, or is it reasonable for platform owners to limit third-party access?
These questions have no easy answers. They involve tradeoffs between integration efficiency and competitive diversity, between platform owners’ property rights and users’ interests in choice, and between innovation incentives and competitive fairness.
The investigation’s outcome will indicate where Italian and EU authorities draw these lines, at least for now. However, as technology evolves, these questions will require continuous reassessment.
Conclusion
Italy’s expanded investigation into Meta’s WhatsApp AI integration and business policy restrictions represents far more than a routine competition enforcement action. It constitutes a critical test of how existing competition law applies to the emerging AI landscape, where platform dominance, data advantages, and distribution control could determine market structure for years to come.
The AGCM’s decision to broaden the probe while initiating interim measures procedures signals that regulators view Meta’s conduct as potentially serious enough to warrant urgent intervention. With WhatsApp’s 37 million Italian users, Meta’s $164.5 billion in global revenue, and potential fines reaching 10% of worldwide turnover, the financial and strategic stakes are enormous.
For Meta, the investigation presents both immediate challenges and longer-term strategic questions. Can the company’s technical justifications for restricting third-party AI access withstand regulatory scrutiny? Will courts and authorities accept that AI integration represents legitimate product improvement rather than anticompetitive exclusion? How should Meta balance its innovation ambitions with its responsibilities as a dominant platform operator?
For competing AI services, the outcome determines whether they can access one of the world’s most important distribution channels or must rely exclusively on standalone apps and alternative platforms. For users, the investigation examines whether they benefit from integrated convenience or suffer from reduced choice.
For regulators, the case tests whether traditional competition law tools designed for industrial-era markets can effectively govern fast-moving digital platform ecosystems. It explores whether authorities can intervene quickly enough to prevent irreversible competitive harm while maintaining rigorous evidentiary standards.
As the investigation proceeds toward its scheduled December 31, 2026, conclusion, with the critical January 15, 2026, deadline for third-party AI service termination approaching, the technology industry, legal community, and millions of WhatsApp users will watch closely. The outcome will establish precedents shaping AI competition, platform regulation, and digital market governance across Europe and potentially globally.
One certainty emerges from this complex situation: the intersection of artificial intelligence, platform dominance, and competition law will generate increasing regulatory attention and enforcement action. The Italian WhatsApp investigation represents not an isolated incident but rather the opening chapter in what promises to be a prolonged period of intensive scrutiny on how dominant technology platforms integrate AI capabilities while managing their competitive responsibilities.
Sources
- MarketScreener. (2025, November 26). Italy’s Competition Watchdog Broadens WhatsApp AI Policy Probe. https://www.marketscreener.com/news/italy-s-competition-watchdog-broadens-whatsapp-ai-policy-probe-ce7d5ed2d18ef122
- The Business Standard. (2025, November 26). Italy antitrust watchdog may curb Meta as WhatsApp AI probe widens. https://www.tbsnews.net/world/italy-antitrust-watchdog-may-curb-meta-whatsapp-ai-probe-widens-1295581
- Investing.com. (2025, November 26). Italy competition watchdog broadens probe into Meta over AI tools in WhatsApp. https://www.investing.com/news/stock-market-news/italy-competition-watchdog-broadens-probe-into-meta-over-ai-tools-in-whatsapp-4378831
- Investing.com. (2025, November 26). Italian antitrust watchdog widens probe into Meta over WhatsApp AI tools. https://www.investing.com/news/stock-market-news/italian-antitrust-watchdog-widens-probe-into-meta-over-whatsapp-ai-tools-4379467
- GuruFocus. (2025, November 26). Meta Faces Expanded Italian Antitrust Probe Over WhatsApp AI Tools. https://www.gurufocus.com/news/3225237/meta-faces-expanded-italian-antitrust-probe-over-whatsapp-ai-tools
- All About AI. (2025, July 30). Meta faces Italian scrutiny over AI tool integration in WhatsApp. https://www.allaboutai.com/ai-news/meta-faces-italian-scrutiny-over-ai-tool-integration-in-whatsapp/
- Digital Watch Observatory. (2025, July 31). Italy investigates Meta over AI integration in WhatsApp. https://dig.watch/updates/italy-investigates-meta-over-ai-integration-in-whatsapp
- Techpoint Africa. (2025, August 1). I tested Meta AI on WhatsApp (2025): how it works & how to use it. https://techpoint.africa/guide/meta-ai-on-whatsapp-how-to-use/
- Data Studios. (2024). Meta AI in WhatsApp: Assistant Behavior, Model Updates, and Data Privacy Controls. https://www.datastudios.org/post/meta-ai-in-whatsapp-assistant-behavior-model-updates-and-data-privacy-controls
- TechCrunch. (2025, June 25). Meta is adding AI-powered summaries to WhatsApp. https://techcrunch.com/2025/06/25/meta-is-adding-ai-powered-summaries-to-whatsapp/
- The Hacker News. (2025, June 26). WhatsApp Adds AI-Powered Message Summaries for Faster Chat Previews. https://thehackernews.com/2025/06/whatsapp-adds-ai-powered-message.html
- Chatarmin. Meta AI WhatsApp – Features, Tips & Insights on the New AI Assistant. https://chatarmin.com/en/blog/meta-ai-whats-app
- Dataconomy. (2025, March 21). Meta AI Just Landed On WhatsApp, Instagram And Messenger In Europe. https://dataconomy.com/2025/03/21/meta-ai-just-landed-on-whatsapp-instagram-and-messenger-in-europe/
- Kluwer Competition Law Blog. Did Meta Tie its AI Assistant to WhatsApp? https://legalblogs.wolterskluwer.com/competition-blog/did-meta-tie-its-ai-assistant-to-whatsapp/
- Smallest.ai. WhatsApp Beta Integrates AI-Suggested Topics and Voice Chat. https://smallest.ai/blog/whatsapp-beta-integrates-ai-suggested-topics-and-voice-chat-a-step-toward-smarter-messaging
- Meta AI. https://ai.meta.com/meta-ai/
- Cambridge University Press. Article 102 TFEU – abuse of a dominant position. https://www.cambridge.org/core/books/abs/an-introduction-to-eu-competition-law/article-102-tfeu-abuse-of-a-dominant-position/FB90CF417C50B1ED22D7808F5E24EBE2
- Wikipedia. (2025, October 24). Article 102 of the Treaty on the Functioning of the European Union. https://en.wikipedia.org/wiki/Article_102_of_the_Treaty_on_the_Functioning_of_the_European_Union
- EUR-Lex. Consolidated version of the Treaty on the Functioning of the European Union. https://eur-lex.europa.eu/legal-content/EN/TXT/HTML/?uri=CELEX:12008E102
- Global Competition Review. European Union: Abuse of dominance and article 102 of the TFEU. https://globalcompetitionreview.com/review/the-european-middle-east-and-african-antitrust-review/2023/article/european-union-abuse-of-dominance-and-article-102-of-the-tfeu
- European Commission. Article 102 Investigations – Competition Policy. https://competition-policy.ec.europa.eu/antitrust-and-cartels/procedures/article-102-investigations_en
- European Commission. Application of Article 102 TFEU. https://competition-policy.ec.europa.eu/antitrust-and-cartels/legislation/application-article-102-tfeu_en
- Global Competition Review. (2025). European Union: Article 102 TFEU and the European Commission’s 2025 Guidelines. https://globalcompetitionreview.com/review/the-european-middle-east-and-african-antitrust-review/2025/article/european-union-article-102-tfeu-and-the-european-commissions-2025-guidelines-what-expect-and-what-ask
- Oxford Academic. (2024, June 7). Critical Inquiry into ‘Abuse’ in EU Competition Law. https://academic.oup.com/ojls/article/44/2/405/7624134
- DoubleClick. (2024, June 7). WhatsApp User Statistics for 2024. https://doubletick.io/blog/whatsapp-user/
- World Population Review. WhatsApp Users by Country 2025. https://worldpopulationreview.com/country-rankings/whatsapp-users-by-country
- WhatsTheBigData. (2024, January 1). WhatsApp Statistics, Users, Demographics as of 2024. https://whatsthebigdata.com/whatsapp-statistics/
- Skillademia. (2025, April 7). WhatsApp Statistics (2025): Number of Downloads, Demographics, and More. https://www.skillademia.com/statistics/whatsapp-statistics/
- Business of Apps. WhatsApp Revenue and Usage Statistics (2025). https://www.businessofapps.com/data/whatsapp-statistics/
- WANotifier. WhatsApp Statistics 2025: Usage Trends, Demographics & More. https://wanotifier.com/whatsapp-statistics/
- GrabOn. (2025, July 4). WhatsApp Statistics 2025: Users, Market Share & Revenue. https://grabon.com/blog/whatsapp-statistics/
- SignHouse. (2024, August 1). WhatsApp Revenue and Growth Statistics (2024). https://usesignhouse.com/blog/whatsapp-stats/
- ThunderBit. 100 WhatsApp Statistics And Fact. https://thunderbit.com/blog/whatsapp-statistics
- WhatsApp. WhatsApp Business Solution Terms. https://www.whatsapp.com/legal/business-solution-terms
- SleekFlow. (2021, October 25). WhatsApp Business API: A definitive guide for your business. https://sleekflow.io/blog/whatsapp-business-api
- WhatsApp. Meta Terms for WhatsApp Business. https://www.whatsapp.com/legal/meta-terms-whatsapp-business
- Lime Technologies. (2024, March 14). WhatsApp Business API – Everything You Need to Know [2024]. https://connect.lime-technologies.com/en/blog/whatsapp-api/
- Windows Forum. (2025, November 26). Copilot Leaves WhatsApp by Jan 15 2026: Policy Change. https://windowsforum.com/threads/copilot-leaves-whatsapp-by-jan-15-2026-policy-change.391221/
- Yellow.ai. (2025, July 11). What is WhatsApp business API? The Complete guide (2024). https://yellow.ai/blog/whatsapp-business-api/
- Business Chat. WhatsApp API: the ultimate guide to WhatsApp Business API (2024). https://www.businesschat.io/post/whatsapp-business-platform-ultimate-guide
- WANotifier. Ultimate Guide to WhatsApp API for Businesses (2025). https://wanotifier.com/whatsapp-business-api-guide-for-businesses/
- Windows Forum. (2025, November 28). WhatsApp API Bans Third Party AI; Meta AI Becomes In App Default. https://windowsforum.com/threads/whatsapp-api-bans-third-party-ai-meta-ai-becomes-in-app-default.391449/
- Windows Forum. (2025, November 29). WhatsApp Policy Bans General AI Bots as Copilot Leaves Jan 15 2026. https://windowsforum.com/threads/whatsapp-policy-bans-general-ai-bots-as-copilot-leaves-jan-15-2026.391495/
- NewsBytes. (2025, November 26). Why Italy is expanding its antitrust probe into Meta. https://www.newsbytesapp.com/news/business/italy-expands-antitrust-investigation-into-meta-s-ai-tools-on-whatsapp/story
- WinBuzzer. (2025, July 30). Italy Raids Meta Offices and Launches Antitrust Probe Over ‘Imposed’ WhatsApp AI Chatbot. https://winbuzzer.com/2025/07/30/italian-raids-meta-offices-and-launches-antitrust-probe-over-imposed-whatsapp-ai-chatbot-xcxwbn/
- Cryptopolitan. (2025, July 30). Meta faces antitrust probe in Italy over AI integration in WhatsApp. https://www.cryptopolitan.com/meta-faces-antitrust-probe-in-italy/
- CryptoRank. Italy’s antitrust watchdog (AGCM) is investigating Meta. https://cryptorank.io/news/feed/621b3-meta-faces-antitrust-probe-in-italy
- Global Banking and Finance. (2025, July 30). Meta faces Italian competition investigation over WhatsApp AI chatbot. https://www.globalbankingandfinance.com/ITALY-ANTITRUST-f89647a6-c5f0-48bf-add4-4c377c95ce64
- Mitrade. (2025, July 30). Italy’s antitrust watchdog (AGCM) is investigating Meta. https://www.mitrade.com/au/insights/news/live-news/article-3-998726-20250731
- Bitcoin Ethereum News. (2025, July 30). Italy’s antitrust watchdog (AGCM) is investigating Meta. https://bitcoinethereumnews.com/finance/italys-antitrust-watchdog-agcm-is-investigating-meta/
- Yahoo Finance. (2025, July 30). Meta Faces Italy Probe Over WhatsApp AI. https://uk.finance.yahoo.com/news/meta-faces-italy-probe-over-231019408.html
- Storyboard18. (2025, July 30). Meta under Italian antitrust investigation for integrating AI chatbot into WhatsApp. https://www.storyboard18.com/brand-marketing/meta-under-italian-antitrust-investigation-for-integrating-ai-chatbot-into-whatsapp-77884.htm
- Meta Platforms. (2025, January 29). Meta Reports Fourth Quarter and Full Year 2024 Results. https://investor.atmeta.com/investor-news/press-release-details/2025/Meta-Reports-Fourth-Quarter-and-Full-Year-2024-Results/
- MacroTrends. Meta Platforms Revenue 2011-2025. https://www.macrotrends.net/stocks/charts/META/meta-platforms/revenue
- PR Newswire. (2025, January 29). Meta Reports Fourth Quarter and Full Year 2024 Results. https://www.prnewswire.com/news-releases/meta-reports-fourth-quarter-and-full-year-2024-results-302363791.html
- Variety. (2024, October 30). Meta Q3 earnings report 2024. https://variety.com/2024/digital/news/meta-q3-2024-results-record-revenue-1236195075/
- Stock Analysis. Meta Platforms (META) Revenue 2010-2025. https://stockanalysis.com/stocks/meta/revenue/
- Voronoi. Meta’s Revenue Breakdown in 2024. https://www.voronoiapp.com/technology/Metas-Revenue-Breakdown-in-2024–4415
- Yahoo Finance. (2025, January 31). Meta Platforms Full Year 2024 Earnings: EPS Beats Expectations. https://finance.yahoo.com/news/meta-platforms-full-2024-earnings-105142090.html
- Visual Capitalist. (2025, March 21). Charted: How Does Meta Make Money? https://www.visualcapitalist.com/charted-how-does-meta-make-money/
- SEC. Meta Platforms Inc. Form 10-K for fiscal year ending December 31, 2024. https://www.sec.gov/Archives/edgar/data/1326801/000132680125000017/meta-20241231.htm
- CNBC. (2024, October 30). Meta Q3 earnings report 2024. https://www.cnbc.com/2024/10/30/meta-q3-earnings-report-2024.html
- Windows Forum. (2025, November 27). Italy Probes Meta WhatsApp AI Policy as Interim Measures Loom. https://windowsforum.com/threads/italy-probes-meta-whatsapp-ai-policy-as-interim-measures-loom.391356/
- Agenda Digitale. (2025, November 27). Agcm su Meta AI, perché è un caso da manuale per il futuro. https://www.agendadigitale.eu/mercati-digitali/agcm-su-meta-ai-perche-e-un-caso-da-manuale-per-il-futuro/
- The Legal Wire. (2025, July 29). Italy Competition Authority investigates into Meta’s integration of Meta AI in WhatsApp. https://thelegalwire.ai/italy-competition-authority-investigates-into-metas-integration-of-meta-ai-in-whatsapp/
- Yahoo Finance. (2025, July 31). Italy’s AGCM probes Meta’s WhatsApp AI integration. https://finance.yahoo.com/news/italy-agcm-probes-meta-whatsapp-092012177.html
- Digital Policy Alert. Competition authority opened investigation into Meta for alleged abuse of dominant position. https://digitalpolicyalert.org/event/32290-competition-authority-opened-investigation-into-meta-for-alleged-abuse-of-dominant-position-meta-platforms-inc-et-al-a576
- US News. (2025, November 26). Italy Antitrust Watchdog May Curb Meta as WhatsApp AI Probe Widens. https://money.usnews.com/investing/news/articles/2025-11-26/italy-competition-watchdog-broadens-probe-into-meta-over-ai-tools-in-whatsapp
- The Star. (2025, November 26). Italy antitrust watchdog may curb Meta as WhatsApp AI probe widens. https://www.thestar.com.my/tech/tech-news/2025/11/26/italy-broadens-antitrust-probe-into-meta-over-ai-tools-in-whatsapp
- WinBuzzer. (2025, November 26). WhatsApp: Italy to Block Competitor AI Ban with Emergency Antitrust Order. https://winbuzzer.com/2025/11/26/italy-moves-to-block-whatsapps-competitor-ai-ban-with-emergency-antitrust-order-xcxwbn/
