OpenAI Turns ChatGPT Into a Full-Fledged Digital Assistant

Crypto News AI

Written by Aaron S.,

Editor-In-Chief

Last Updated: July 18, 2025

Key Takeaways

OpenAI’s agent mode lets ChatGPT schedule, create slides, and run code from simple prompts;
The agent links to apps like Gmail, uses a terminal, and handles multi-step tasks automatically;
Added safeguards and a no-memory feature help reduce risks in sensitive or harmful use cases.

Stop overpaying - start transferring money with Ogvio. Sign up, invite friends & grab Rewards now! 🎁

OpenAI Turns ChatGPT Into a Full-Fledged Digital Assistant

BITDEGREE IN YOUR SOCIAL FEED

OpenAI has introduced a new "agent mode" for ChatGPT, available to users on Pro, Plus, and Team plans starting July 17.

This mode turns the chatbot into a versatile assistant that can schedule events, build presentations, and write and run code, all by following natural-language prompts.

The agent merges Operator’s ability to interact with websites and Deep Research’s skill at summarizing information from multiple sources.

Is Your Crypto Safe? (5 Best Crypto Security Practices Explained)

Did you know?

Want to get smarter & wealthier with crypto?

Subscribe - We publish new crypto explainer videos every week!

Users can also link the agent to external apps like Gmail and GitHub through connectors. It has access to a built-in terminal and can use app interfaces to complete specific actions.

These tools enable the agent to manage detailed requests, such as planning and shopping for a meal, comparing competitors, and preparing a slide deck. The assistant carries out the necessary steps automatically, based only on the prompt given.

OpenAI has implemented additional safeguards for the agent mode. The company marks the agent as high risk in sensitive areas, such as biology and chemistry. Even though it has not seen harmful use yet, it checks all prompts for biological content and screens any related responses to prevent misuse.

Furthermore, OpenAI has disabled the memory feature in agent mode. While memory in other parts of ChatGPT lets it reference earlier conversations, the company stated that leaving it on could allow attackers to extract sensitive information.

Recently, Microsoft introduced a new AI system, which it claims can outperform doctors when diagnosing complex medical cases. How does it work? Read the full story.