Meta Platforms, Inc. (META.US) announced on Wednesday the launch of a new feature called the Meta Business Agent. This AI agent is designed for businesses of all sizes, capable of answering customer queries, recommending products, scheduling services, and screening and identifying potential sales leads. The agent can also search through chat histories and compile a "morning briefing," providing businesses with key insights and important information from relevant conversation threads. In the future, the agent's capabilities will be expanded to include tasks such as conducting market research, analyzing product insights, and providing "competitive intelligence."
Meta CEO Mark Zuckerberg stated, "Today I'm officially launching the Meta Business Agent to help businesses large and small equip themselves with their own dedicated agent to connect with customers and assist in daily operations. Now, a clothing boutique in Birmingham or a bakery in São Paulo can offer highly personalized, 24/7 customer service on par with major brands."
The Meta Business Agent is currently being rolled out initially to a select group of businesses across Meta's suite of applications, including WhatsApp, Messenger, and Instagram. For years, WhatsApp has served as a vital communication platform for businesses globally, regardless of their size. Meta is now integrating artificial intelligence into it, aiming to transform WhatsApp into a practical workflow software solution for small and medium-sized enterprises.
Meta will introduce a search engine function within WhatsApp powered by the Business Agent. Users will be able to more easily search for and discover businesses using information such as company names, phone numbers, or contact names. According to a press release from Meta, over one million businesses are already using the Meta Business Agent on WhatsApp and Messenger to respond to customer needs around the clock.
Additionally, Meta has launched the Meta Business Agent Platform. This is a platform for building, customizing, and deploying business agents at scale. Some analysis suggests that the launch of the Business Agent represents an attempt by Meta to open up a significant new revenue stream for its social applications. By the fourth quarter of 2025, the annualized revenue generated from businesses paying to communicate with customers via WhatsApp had already surpassed $2 billion. However, this still represents only a small fraction compared to Meta's total revenue of approximately $200 billion last year.
Meta plans to include this agent in some of its WhatsApp Business Premium subscription packages and will charge businesses for using the agent. The company indicated that larger enterprises will be charged based on the volume of tokens they use. Meta is seeking to reduce its dependence on advertising revenue, leveraging artificial intelligence (AI) to develop new income sources, and setting its sights on the booming AI agent market.
Since Facebook launched its targeted user advertising business in 2007, Zuckerberg has built a mature profit system based on online ads. Currently, nearly 98% of Meta's revenue still comes from advertising, and the company's attempts over the years in digital products and physical goods sales have repeatedly fallen short. While AI has helped solidify its core advertising business, Zuckerberg harbors a larger vision for AI, aiming to compete with OpenAI, Anthropic, and Google in the markets for cutting-edge large language model development and high-level commercial services.
Over the past year, Zuckerberg has significantly adjusted Meta's AI strategy, investing heavily to poach top talent from competitors like OpenAI and Google, and in April of this year, released the Muse Spark model. Zuckerberg stated that Meta is currently "building agent capabilities" that can perform advanced operational tasks, such as providing strategies for business growth and customer acquisition, compiling industry competitor intelligence, and conducting real-time reviews of business strengths and weaknesses. He said, "As our self-developed large models continue to evolve, the range of tasks agents can undertake will grow, ultimately enabling them to fully manage the entire operational chain of a business."