Meta Faces Antitrust Probe Over WhatsApp API
The COMESA Competition Authority in Africa is investigating Meta for potentially anti-competitive practices related to its WhatsApp Business API. The probe focuses on allegations that Meta may be restricting third-party AI chatbot integrations to favor its own tools. This regulatory action could signal similar scrutiny in other emerging markets, including India, where WhatsApp is a dominant platform for small business digitization.
- The COMESA Competition and Consumer Commission's investigation began on February 17, 2026, following a complaint by a Uganda-based non-profit, AdLegal, on January 5, 2026. The probe centers on whether Meta's October 15, 2025, amendment to its WhatsApp Business Solution Terms constitutes an abuse of a dominant market position. - Meta's updated policy, which took full effect on January 15, 2026, explicitly prohibits "general purpose" AI chatbots from using the WhatsApp Business API. This directly impacts companies like OpenAI, Perplexity AI, and Luzia, while leaving Meta AI as the sole general-purpose AI assistant on the platform. - The policy distinguishes between general-purpose AI and business-specific AI, still permitting AI-powered chatbots for defined commercial functions like customer support, order tracking, and sales. Meta has stated the change is to prevent infrastructure strain and refocus the API on its core purpose of business-to-customer communication. - This investigation is part of a global pattern of regulatory scrutiny over this specific policy change, with similar antitrust probes underway by the European Union, Italy, and Brazil. In December 2025, the Italian competition authority ordered Meta to halt the ban pending its investigation. - The complaint alleges that by blocking competing AI products like Google Gemini and OpenAI's ChatGPT from the API, while giving its own Meta AI full access, Meta is restricting market entry and foreclosing competition. - Meta has defended its decision, citing the need to protect the integrity of the platform and arguing that its systems were not originally designed to support the high volume of interactions generated by general-purpose AI chatbots.
