OpenAI Boosts Operator with o3 for Enhanced Reasoning and Safety

Published Date : 25/05/2025 

OpenAI is rolling out an advanced AI model, o3, to power its Operator, enhancing its capabilities in reasoning and safety. This upgrade is part of the ongoing competition among AI firms to develop reliable digital agents. 

OpenAI, a leading artificial intelligence (AI) company, is updating the AI model that powers Operator, its autonomous AI agent. Operator can browse the web and interact with software in a cloud-hosted virtual machine to perform user requests. The new model, o3, is part of OpenAI’s o series of reasoning models, replacing the previous GPT-4o-based model.

By several benchmarks, o3 is a more advanced model, particularly on tasks requiring mathematical ability and reasoning. “We are replacing the existing GPT-4o-based model for Operator with a version based on OpenAI o3,” OpenAI wrote in a blog post. “The API version (of Operator) will remain based on 4o.”

Operator is part of a growing set of agentic tools developed by AI firms as they compete to build agents capable of performing digital tasks with minimal supervision. Google offers a similar agent through its Gemini API, which can browse the web and take actions on users’ behalf. It also offers a consumer-facing version called Mariner. Anthropic’s models can perform various computer tasks as well, including opening files and navigating webpages.

According to OpenAI, the upgraded Operator model, dubbed o3 Operator, was “fine-tuned with additional safety data for computer use,” using datasets designed to “teach the model (OpenAI’s) decision boundaries on confirmations and refusals.” The company has released a technical report detailing o3 Operator’s performance in safety evaluations. Compared to the GPT-4o version, the new model is less likely to carry out illicit activities, search for sensitive personal data, or fall prey to prompt injection, a common AI attack technique.

“o3 Operator uses the same multi-layered approach to safety that we used for the 4o version of Operator,” OpenAI wrote in its blog post. “Although o3 Operator inherits o3’s coding capabilities, it does not have native access to a coding environment or terminal.”

This upgrade is a significant step forward in the development of AI agents that can reliably perform tasks with minimal human intervention, enhancing both efficiency and safety in the digital realm. 

Frequently Asked Questions (FAQS): 

Q: What is OpenAI's Operator?

A: OpenAI's Operator is an AI agent that can autonomously browse the web and interact with software in a cloud-hosted virtual machine to carry out user requests.

Q: What is the o3 model?

A: The o3 model is a more advanced AI model from OpenAI, particularly effective in tasks requiring mathematical ability and reasoning.

Q: How does the o3 model improve Operator?

A: The o3 model enhances Operator's reasoning capabilities and includes additional safety features, making it less likely to engage in illicit activities or fall prey to AI attacks.

Q: What is the difference between o3 Operator and GPT-4o?

A: o3 Operator is based on the o3 model, which is more advanced in reasoning and safety compared to the GPT-4o model, which was previously used to power Operator.

Q: Are there similar AI agents from other companies?

A: Yes, Google offers a similar agent through its Gemini API, and Anthropic’s models can also perform various computer tasks, including opening files and navigating webpages. 

More Related Topics :