Meta’s business AI has quietly crossed an important threshold: it is now handling 10 million customer conversations every week. That may sound like just another big tech metric, but it signals something deeper. Meta isn’t merely chasing the next flashy foundation model; it’s wiring AI into the everyday workflows of tens of thousands of small businesses across WhatsApp, Instagram and Messenger. In practice, this could shift who controls the relationship between small companies and their customers — and who earns money every time a message is sent.
The news in brief
According to TechCrunch’s report on Meta’s Q1 2026 earnings call, Meta says its business AI tools now facilitate about 10 million conversations per week, up from around 1 million per week at the start of 2026.
This growth follows an expanded beta rollout of Meta’s business AI assistant across the U.S., EMEA, APAC and LATAM. For now, these AI tools are free for most small businesses, but CEO Mark Zuckerberg indicated the company plans to introduce a long‑term monetization model.
Meta is powering these products with its new large language model Muse Spark, developed under the Meta Superintelligence Labs division. On the advertising side, Meta claims over 8 million advertisers used at least one of its generative ad creative tools in the quarter, with tests showing around 3% higher conversion rates for advertisers using AI‑generated video.
The company is also launching an open beta of Meta Ads AI Connectors, which link advertisers’ Meta ad accounts to AI agents. Overall, Meta reported $56.3 billion in revenue and $26.8 billion profit for the quarter, with $885 million from app revenue such as paid messaging on WhatsApp and app subscriptions.
Why this matters
The headline number — 10 million weekly conversations — is strategically more important than it looks. It means Meta is rapidly becoming the default AI front desk for small businesses worldwide.
Who benefits first?
- Small and medium‑sized businesses (SMBs) that could never afford 24/7 support now get automated replies, lead qualification and basic sales assistance for free.
- Meta gains something much more valuable than short‑term revenue: deep integration into everyday business processes and a constant stream of conversational data to improve its models and ad targeting.
Who should be worried?
- Independent chatbot vendors and SaaS helpdesk tools that rely on selling AI support to SMBs. If a “good enough” bot is bundled into WhatsApp or Instagram Business for free, many merchants will never look elsewhere.
- Brands and regulators concerned about data control. Every AI‑mediated interaction sits on Meta’s infrastructure, not the merchant’s.
There’s also a familiar pattern here. Historically, Meta has offered powerful tools to businesses for little or no cost, only to turn on the monetization dial later once dependency is established — think Facebook Pages’ organic reach in the early 2010s, followed by a strong push toward paid promotion.
Today, conversations are free. Tomorrow, it’s easy to imagine charges per AI‑handled message, premium automation tiers or pay‑to‑prioritize replies. For millions of businesses who now rely on WhatsApp and Instagram as lifelines, this could quietly turn into a new operating cost.
The bigger picture
Meta’s move sits at the intersection of three major trends:
Conversational commerce as the default for small business
In emerging markets — and increasingly in Europe — many businesses skipped the website era and went straight to WhatsApp‑first or Instagram‑first operations. AI inside these channels doesn’t just add convenience; it formalizes that model. Instead of a helpdesk plus a CRM plus a webshop, you get an AI layer sitting on top of your DMs.AI‑native advertising and creative
With 8 million advertisers already using generative tools, Meta is accelerating the shift from manual creative production to continuous, AI‑driven experimentation. This echoes what Google has done with Performance Max and what Amazon offers with automated creatives, but Meta has a key advantage: it owns the social and messaging context where those ads appear and where conversations continue.The rise of AI “middleware”
The new Meta Ads AI Connectors are more than a convenience feature. They’re a bridge between brands’ own AI agents and Meta’s advertising and messaging stack. Over time, this could turn Meta into essential middleware for any company wanting to run an AI agent that both talks to customers and buys ads on their behalf.
Historically, whenever Meta has controlled the discovery and engagement layer — News Feed before, Reels more recently — it has eventually squeezed that control into ad dollars. Business AI is the logical extension into post‑click interactions: everything that happens after the user taps an ad or message.
In a world where OpenAI, Google and others dominate the narrative about frontier models, Meta is quietly building something harder to replicate: distribution plus habit. Users are already on WhatsApp and Instagram. Now the AI shows up where they already spend their time.
The European and regional angle
For European users and companies, this development collides directly with the EU’s regulatory stack: GDPR, the Digital Services Act (DSA), the Digital Markets Act (DMA) and the new EU AI Act.
Meta is officially a DMA “gatekeeper”, which means any deep integration of AI into business messaging will face questions about self‑preferencing and lock‑in. If Meta’s own AI assistant is promoted more aggressively than third‑party tools, expect scrutiny from Brussels and national competition authorities.
Under GDPR, the data flowing through these AI conversations — often sensitive customer details, complaints, purchase intent — raises questions about joint controllership. Who is responsible if the AI stores or uses data in ways that violate consent? The small merchant, Meta, or both?
The AI Act adds another layer: depending on how these tools are used (e.g. for credit‑related decisions, profiling or employment), they could fall into high‑risk categories, triggering strict transparency, logging and human‑oversight requirements. Many SMEs will not have the legal or technical capacity to understand what exactly the AI is doing with their data.
European markets also differ culturally. In Germany, Austria or the Nordics, where privacy sensitivity is high, customers may push back harder against opaque AI‑driven chats than in markets where WhatsApp is already the de facto business channel. In Southern and Eastern Europe, where WhatsApp and Instagram‑based micro‑businesses are exploding, adoption may be faster — but so will dependence on Meta’s infrastructure.
The opportunity for European tech ecosystems is to build privacy‑preserving, open alternatives: AI‑powered CRMs and messaging platforms that can integrate with, but not be fully subsumed by, Meta’s stack.
Looking ahead
Meta’s playbook is familiar, and you can almost sketch the next steps:
Scale first, monetize later
Expect conversation volumes and advertiser AI usage to be the key metrics Meta highlights in coming quarters. Once growth stabilizes, new pricing models will likely appear: per‑conversation charges (similar to today’s WhatsApp Business API), tiered AI capabilities, or bundled packages tied to ad spend.Deeper integration into workflows
The current assistants mostly handle FAQs and basic sales queries. The next wave will plug into inventory systems, booking engines, and payments. When your AI on WhatsApp can not only answer a question but also modify an order or issue a refund, switching away from Meta becomes much harder.Regulatory flashpoints
Watch for early test cases in the EU where AI‑driven miscommunication leads to consumer harm — wrong prices, discriminatory replies, or misleading product claims. Who carries the legal risk when the AI “improvises”? Regulators will be forced to clarify liability.Competitive responses
Google will lean even harder on its own AI‑driven ads and Business Messages. Shopify, Salesforce and smaller SaaS vendors will pitch themselves as more controllable, less extractive alternatives. But none of them have the same grip on consumer messaging that Meta enjoys.
Over the next 18–24 months, the most important question is not whether Meta can build good enough AI — clearly, it can — but whether regulators and businesses will accept Meta as the operating system for customer conversations.
The bottom line
Meta’s 10 million weekly AI‑mediated conversations are a leading indicator of a bigger shift: customer service and sales for small businesses are moving inside a few giant messaging platforms, with Meta in the lead. The tools are free today, but history suggests the bill will come later. The real question for European businesses and regulators is simple: do we want the front door to our customer relationships controlled by one company’s AI, and on whose terms?
