Meta Platforms has stepped onto the enterprise‑AI stage with a new “business agent” that can handle everything from scheduling appointments to closing sales directly inside its consumer apps. Unveiled at the London‑based Conversations conference, the tool expands the capabilities already offered on WhatsApp and Messenger and will soon appear on Instagram, signalling Meta’s intent to challenge AI powerhouses such as OpenAI, Anthropic and Google in the corporate market.
The announcement comes on the back of over a million businesses that currently use earlier versions of Meta’s chatbots on its messaging platforms. By adding “agentic” functions—where the assistant can actually execute transactions, process payments and place orders—Meta hopes to transform casual messenger bots into genuine productivity partners for South African firms of all sizes.
Naomi Gleit, Meta’s head of product, told reporters the rollout is “definitely an enterprise play.” The new business agent can be customised to reflect a brand’s tone, answer frequent‑question‑type enquiries, qualify leads and, when a query becomes too complex, hand it over to a human operator. Initial access will be free, with tiered subscription plans slated for launch later in the year.
Meta business AI agent powers the next generation of enterprise workflows
The business agent platform goes beyond Meta’s in‑app tools, offering a sandbox where developers can build bespoke AI agents that operate across hundreds of third‑party systems, including Shopify, Zendesk and Shopee. Gleit explained that the platform provides “enterprise‑grade controls, guardrails and measurement,” allowing larger organisations to monitor and fine‑tune agent performance.
To accelerate adoption, Meta is forming an “Enterprise Solutions” squad that will embed engineers within client sites—a play borrowed from Anthropic’s forward‑deployed model. These teams will help businesses navigate internal resistance to AI, write custom code and ensure the agents integrate smoothly with existing processes.
Below is a snapshot of how Meta’s offering stacks up against its main rivals in the enterprise‑AI space:
| Feature | Meta Business AI Agent | OpenAI Enterprise | Anthropic Claude+ | Google Cloud Vertex AI |
|---|---|---|---|---|
| Native integration with social apps | WhatsApp, Messenger, Instagram | None | None | Limited (via APIs) |
| Third‑party system connectors | 200+ (Shopify, Zendesk, Shopee…) | 100+ (via partners) | 80+ (custom SDK) | 150+ (Google ecosystem) |
| Free tier | Yes (limited actions) | No | No | Yes (basic) |
| Customisable brand tone | Built‑in UI | API‑only | API‑only | API‑only |
| Enterprise‑grade guardrails | Included | Add‑on | Add‑on | Included |
| On‑site engineering support | Yes (Enterprise Solutions) | No | No | No |
The table highlights Meta’s unique advantage: deep integration with its own consumer platforms and a free entry point that could appeal to SMEs looking for a low‑cost AI boost. While OpenAI and Google offer broader model libraries, they lack the built‑in messaging reach that Meta can immediately leverage.
Gleit stressed that the ambition is to give small businesses a single pane of glass for all their digital tasks. “The number one thing I hear, especially from small businesses, is ‘I just want to go to one place that can do all the things’,” she said. “You want to make things modular, and you also need to be willing to evolve, because the technology is moving so quickly.” This modularity is reflected in the platform’s ability to plug into external services while still operating inside familiar chat environments.
Nevertheless, the rollout is not without risk. Earlier this week an incident surfaced where hackers manipulated Meta’s AI support chatbot into divulging access to high‑profile Instagram accounts. Gleit clarified that the breach stemmed from a faulty technical check rather than the agent itself, noting, “The agent actually exposed a technical check that wasn’t working.” Meta is investigating the flaw and promises tighter safeguards before the business agents gain broader transactional authority.
For South African enterprises, the implications are significant. Companies that already run advertising and customer service through Meta’s ad suite can now centralise sales, bookings and support within the same ecosystem, potentially reducing reliance on disparate SaaS tools. Moreover, the free tier lowers the barrier for startups in Cape Town’s tech hub or Durban’s manufacturing SMEs to experiment with AI‑driven workflows without immediate capital outlay.
Industry observers suggest the move could shift how local advertisers allocate budgets. By bundling AI agents with ad products, Meta may entice businesses to consolidate spend on its platforms, echoing trends seen in global markets where AI‑enhanced ad placements drive higher ROI. As the platform matures, integration with South Africa’s e‑commerce players—such as Takealot and Superbalist—could further cement Meta’s foothold in the digital commerce chain.
The emerging business agent platform also opens doors for custom internal tools, from inventory management bots to HR assistants. Meta’s plan to offer “agentic AI products” for internal functions signals a broader strategy to embed AI across the entire corporate stack, not just customer‑facing channels.
Overall, Meta’s foray into enterprise AI appears aimed at leveraging its massive user base to create a seamless bridge between social interaction and business operations. By offering a free, highly integrated agent now and rolling out paid tiers with advanced controls later, the company is positioning itself as a one‑stop shop for South African firms eager to automate and scale. The next months will reveal whether the platform’s promise holds up under real‑world demand and whether the technical hiccups can be ironed out before the agents start processing payments on behalf of local businesses.