USA (Commonwealth Union)_ OpenAI has just taken a major step forward in making artificial intelligence more useful in everyday life. Accordingly, the company recently unveiled the ChatGPT Agent, an advanced AI tool designed to handle complex digital tasks on its own, and one of the key minds behind this innovation is Yash Kumar, an engineer of Indian origin who is helping shape the future of AI from inside OpenAI’s San Francisco headquarters. In a recent demonstration, Kumar, along with OpenAI research lead Isa Fulford, gave the world a first look at what the ChatGPT agent can actually do. Unlike previous versions of ChatGPT that mainly assisted through conversation or basic tasks, this new agent is capable of doing much more, including operating a virtual computer system to complete tasks without constant user input.
Also read: They Came with Dreams — Now These 10 Indian-Americans Control Billions
Kumar explained that the ChatGPT Agent combines two of OpenAI’s earlier experimental projects, Operator, which allowed AI to navigate the web and perform actions, and Deep Research, a tool that could analyze large volumes of information. Now, both capabilities are merged into one powerful system that can not only find answers but also complete multi-step jobs for users. That includes everything from browsing websites and generating reports to creating slideshows, organizing calendars, running code, and managing spreadsheets, all within a secure, private computing environment.
Originally from India, Yash Kumar is a graduate of the International Institute of Information Technology (IIIT) Hyderabad, one of India’s leading computer science institutions. After earning his degree, he joined Google, where he worked as a software engineer for eight years. He then took on leadership roles at companies like Scratch and DoorDash, where he served as Head of Engineering and Director, respectively. He joined OpenAI in late 2023 and now plays a critical role in developing some of the organization’s most advanced AI tools. In the recent media session, Kumar emphasized that the agent doesn’t just “chat,” it takes real action. “It has access to an entire computer,” he said. This means that the Agent can carry out complete tasks from beginning to end, while users maintain control throughout the process. The AI asks for approval before doing anything sensitive, like sending emails or booking appointments.
Also read: As Scandal Explodes in South African Police, Ramaphosa Turns to a Trusted Veteran
The ChatGPT Agent is meant to act like a helpful digital assistant that doesn’t need constant instructions. If you start a task, say, planning a trip or building a presentation, the Agent can carry it forward independently. It even checks in with you along the way if it needs more details or if you want to adjust its approach. OpenAI also added features that allow users to pause a task, get a status update, or even stop it entirely and still keep partial results. And if you’re using the ChatGPT app on your phone, it’ll notify you once your task is complete. According to OpenAI, this tool was designed with security in mind. Even though the system is powerful, it has built-in safety measures, including protections originally created to prevent misuse in sensitive fields like biology and chemistry. There’s no indication that the ChatGPT Agent poses any such risks, but the company says it’s staying cautious.
Also read: Jeff Williams Is Out—Meet the Man Now Steering Apple’s Global Machine
In a post on X, OpenAI CEO Sam Altman described the ChatGPT Agent as a major leap in what AI can do: “It can think for a long time, use tools, take actions, and think some more, all while working toward your goals,” he wrote. Ultimately, the new ChatGPT Agent shows just how far AI has come and how it’s being shaped by global talent like Yash Kumar. With tools like this, OpenAI hopes to turn AI into a true partner in work, planning, and everyday life.
