Meta Platforms CEO Mark Zuckerberg is working on creating a personal agent based on artificial intelligence, which should help him in performing management tasks. This project is part of the company’s strategy aimed at faster implementation of AI in internal processes. It is expected that the new tool will simplify work, help reduce bureaucratic procedures and speed up decision-making, reports The Wall Street Journal.
The development is still ongoing, but this tool is already being used for faster access to information. In particular, the agent is able to provide answers without having to go through multiple management layers.
The initiative is part of Meta’s broader drive to integrate AI into all of its work processes. The company aims to accelerate the pace of work and reduce the number of hierarchical layers to better compete with startups that operate in smaller teams. At the same time, Meta, which employs more than 78,000 people, is reviewing its approach to organizing work amid reports of more than 20% layoffs.
“We are investing in tools designed with artificial intelligence in mind so that Meta employees can achieve greater productivity. We are expanding the capabilities of individual employees and simplifying the structure of teams,” — said Zuckerberg.
AI tools are already widely used within the company. Their use has become one of the indicators of assessing work efficiency. Employees share new ways to use AI and develop their own solutions.
At the same time, employees are encouraged to take training programs and participate in hackathons dedicated to AI. At the same time, some staff are expressing concerns about possible cuts related to automation. Meta has previously carried out large-scale layoffs after rapidly expanding its staff during the pandemic.
Among the company’s internal solutions are personal agents like My Claw, which can work with files, chats, and even interact with other agents. Meta also uses the Second Brain tool, which acts as an “AI chief of staff” and helps with document work. Separately, the company acquired the startup Manus and is implementing its agent technologies into internal processes.
Meanwhile, Meta has already experienced a leak of confidential information due to an erroneous recommendation by an internal AI agent, which gave engineers access to a large amount of user and corporate data for two hours. Experts emphasize that, despite the ability of technologies like Claude Code to perform complex tasks, the lack of long-term context and privacy understanding in AI agents will continue to create new vulnerabilities.
