OpenAI to Launch Autonomous AI Agents Early Next Year

According to reports, OpenAI is developing an autonomous AI agent named "Operator," slated to debut in January 2024 as a research preview and developer tool. This product will have the capacity to independently manage computers and carry out tasks, signaling an escalation in the competition among tech giants within the autonomous AI agent sector.

Recently, Anthropic launched an AI agent with "computer usage" capabilities, and it is reported that Google is also preparing to release a similar product in December of the same year. Although the exact consumer release date for "Operator" has not been announced, its development progress indicates a shift toward AI systems capable of proactively interacting with computer interfaces.

Previously, major AI companies had committed to introducing autonomous AI agents. A few weeks ago, during a Reddit "Ask Me Anything" session, OpenAI CEO Sam Altman stated that while better models are expected in the future, the next significant breakthrough will stem from advancements in agent technology. Additionally, at OpenAI's annual Dev Day media event last month, Chief Product Officer Kevin Weil forecasted that 2025 would be the year when agent systems formally enter the mainstream market.

As AI laboratories face increasing pressure to commercialize their expensive models—especially when incremental improvements may not justify higher pricing to users—autonomous agents are viewed as the next potential breakthrough product. Such an innovation, capable of garnering widespread attention akin to ChatGPT, would substantiate the substantial investments made in AI development.