Open AI, the most famous name in AI since its introduction of Chat GPT two years ago is jumping into the AI agent fray. According to Bloomberg News, Open AI is preparing to launch an agent—currently named “Operator”—that will be able to perform multi-step tasks with minimal supervision.
While generative AI has been widely integrated into automation technology by businesses enabling novel outputs in response to plain language prompts, the LLMs it’s based on have difficulty understanding multi-step prompts and reasoning outside a specific task. AI agents will be able to supercharge those gains due to the ability to understand context, plan workflows, connect to external tools and data, and execute actions that progress toward a defined goal.
According to Bloomberg’s sources, Open AI’s entry into agentic AI the tool is planned to be released in January as a research version and via the company’s API for developers. Open AI is working on several research projects related to agents, Bloomberg reports. The most fully developed is a tool that can perform tasks directly in the browser.