Mark Zuckerberg Is Building an AI Agent to Help Him Serve as CEO

By Long Yue

Source: WallStreetCN

As AI technology becomes more deeply embedded in operations, Meta is attempting to reshape the way work gets done by building an “AI-native” company—and it all starts with CEO Mark Zuckerberg.

Recently, it was reported that Meta CEO Mark Zuckerberg is developing a dedicated “CEO Agent” to help him perform his duties more efficiently.

According to sources familiar with the matter who spoke to The Wall Street Journal, this AI agent is still under development. Its primary function is to help Zuckerberg access information more quickly. In the past, he might have needed to rely on layered reporting structures to obtain answers; now, the AI agent can directly retrieve and deliver the required information to him.

This initiative reflects an internal cultural shift at Meta: accelerating workflow, eliminating redundant organizational layers, and transforming employees’ day-to-day working methods. With approximately 78,000 employees, Meta sees widespread AI adoption as essential to staying competitive against leaner, highly agile AI-native startups.

At Meta’s January earnings call, Zuckerberg previewed AI-driven productivity gains, stating that “one person can do the work of a whole team.” He added: “We’re investing in AI-native tools so individuals at Meta can accomplish more. We’re elevating the role of individual contributors and flattening teams.” He noted that he is already seeing “projects that previously required large teams now being completed by a single highly talented individual.”

Internal AI Adoption: From My Claw to Second Brain

AI tools have rapidly proliferated across Meta. Part of this momentum stems from the fact that AI usage has now become a factor in employee performance evaluations. According to insiders, Meta’s internal message boards are flooded with employees sharing new AI use cases and tools they’ve built using AI.

Employees have begun adopting personal agent tools such as My Claw. These tools can access users’ chat histories and work documents—and even communicate on their behalf with colleagues or with those colleagues’ personal agents.

Another AI tool, named Second Brain, has also gained significant traction internally. Sources say this hybrid tool—somewhere between a chatbot and an agent—was built by a Meta employee using Claude. It indexes and queries documents for projects. In the internal post announcing the tool, the employee described it as “designed to serve as an AI Chief of Staff.”

Even on Meta’s internal message boards, a dedicated group has emerged where employees’ personal agents interact with one another. Meanwhile, Meta recently acquired Moltbook, an AI-agent-focused social media platform, and hired its founder. Meta has also acquired Singapore-based startup Manus—which builds personal agents capable of executing tasks for users—and is currently deploying Manus’ tool internally.

Organizational Restructuring: Hyper-Flat Structures and Layoff Concerns

To accelerate large language model development, Meta recently established a new Applied AI Engineering organization. Reports indicate these teams will adopt hyper-flat structures, with up to 50 individual contributors reporting to a single manager.

Maher Saba, a Meta executive overseeing the new organization, wrote in an internal announcement: “From day one, we designed this organization to be AI-native.” These teams will report to Andrew Bosworth, Meta’s Chief Technology Officer.

However, this rapid transformation—and the heightened focus on AI adoption—has triggered anxiety among some employees about potential layoffs. WallStreetCN recently reported that Meta is planning large-scale layoffs, possibly affecting as much as 20% or more of its workforce. As of the end of last December, Meta employed roughly 79,000 people; a 20% reduction would mean over 15,000 employees cut.