Replit CEO and co-founder Amjad Masad has outlined a vision of the future workplace in which artificial intelligence agents handle a substantial portion of routine business operations while employees focus on decision-making, planning and creative tasks.
Speaking at TechCrunch Disrupt 2026 in San Francisco, Masad described the concept as a "self-driving company", referring to organisations where AI agents manage operational workflows with human oversight.
The remarks come as businesses increasingly experiment with AI agents capable of completing multistep tasks, accessing enterprise systems and assisting employees across software development and other functions.
According to Masad, Replit has already experienced changes in its internal operations as AI agents have become more reliable. The company, which develops AI-powered software creation tools, has integrated autonomous systems into several workplace processes.
He explained that the technology has altered how employees allocate their time, with AI agents undertaking preparatory and repetitive tasks before handing over decisions to human workers.
One example involves Replit's customer support operations. When a support request is received, an AI agent conducts an initial assessment by examining GitHub issues, the company's codebase and relevant tickets on Linear, a project management platform.
The information is then passed to a support employee, who can review the findings and determine the appropriate response. If a software fix is required, another agent can be deployed to assist with implementing the changes.
Masad said this process allows AI systems to complete much of the operational work while employees retain responsibility for decisions requiring judgment.
"You need human judgment; you need taste," Masad said, emphasising that human involvement remains an important part of the workflow.
He also highlighted changes in how Replit's teams collaborate. Employees are spending more time discussing software architecture, evaluating priorities and determining which problems should be addressed, rather than focusing primarily on execution.
The shift reflects a broader change in the use of AI within enterprises, where automated systems are moving beyond generating content or responding to prompts towards completing connected tasks across business applications.
For organisations adopting these technologies, such systems could affect functions including customer service, software engineering, project coordination and internal operations.
However, Masad's description represents Replit's experience and expectations rather than evidence that companies can currently operate without human management.
He also said discussions with other businesses increasingly focus on how AI can improve productivity and make work more engaging by reducing repetitive activities.
The Replit chief's comments come amid continuing industry debate about the effect of AI adoption on employment, workplace responsibilities and organisational structures.
While some companies are exploring automation to improve operational efficiency, questions remain around the extent of human oversight required when AI agents access business systems and carry out tasks.
Masad's proposed model places AI agents within everyday workflows while retaining employees' roles in planning, evaluation and final decision-making.
Disclaimer: This article may include information derived from interviews, press releases, public statements, research, company communications and other publicly available or third-party sources. Such material may be summarised, paraphrased or contextualised for journalistic and editorial purposes. All rights in third-party content remain with their respective owners.