BREAKING NEWS: OpenAI has officially launched the revolutionary ChatGPT Agent, transforming its AI chatbot into a proactive assistant capable of managing tasks independently. This groundbreaking feature, unveiled this week, allows ChatGPT to operate on a virtual computer, executing complex tasks from research to real-time execution without constant user oversight.
This upgrade is set to redefine productivity, bridging the gap between thought and action. Users will now be able to give ChatGPT credentials to access their accounts, including those linked to credit cards and payment methods, effectively allowing the AI to make purchases on their behalf. As this technology evolves, users may find themselves increasingly comfortable with ChatGPT handling financial transactions, establishing spending limits to streamline the process.
Currently, when users instruct ChatGPT Agent to shop online, it can complete everything up to the checkout process. The AI will find the right items and add them to the cart but will pause to seek user permission before finalizing any purchases. This safety feature aims to provide users with peace of mind while navigating the evolving landscape of AI-assisted transactions.
ChatGPT Agent is designed to handle a variety of tasks, from planning meals to analyzing competitors and creating presentations—all while maintaining context throughout the process. The AI seamlessly interacts with websites, filling out forms, clicking links, and managing logins, ensuring a user-friendly experience. For swift, reasoning-focused queries, a streamlined text browser fetches and filters web data efficiently.
Additionally, ChatGPT Agent can execute commands in a virtual terminal, enabling users to manage files and run code directly. It connects to various APIs and tools, such as Gmail and GitHub, to pull personal data and automate actions across platforms. This adaptability allows the AI to choose the best tools for each task while preserving a history of actions for accurate results.
Users are encouraged to provide feedback, interrupt tasks, or steer the AI as needed, with options for notifications upon task completion via the mobile app. The tool can generate spreadsheets, slides, and other files ready for user edits, demonstrating its beta support for presentation creation.
Safety is a paramount concern for OpenAI, which has implemented strict safeguards against potential risks, such as prompt injection. Real-world actions by the AI require explicit user approval, ensuring responsible use of this powerful tool. OpenAI has classified ChatGPT Agent as having high capability in biological and chemical domains under their Preparedness Framework, utilizing advanced refusal training and monitoring.
For users in Australia, the ChatGPT Agent is currently rolling out to paid subscribers, with full access expected to expand shortly to regions outside the restricted EEA. Pro users will receive 400 messages per month, while Plus and Team tiers will offer 40 messages, with additional credits available for purchase. Pricing begins at approximately A$31 per month for Plus, A$308 for Pro, and A$38 per user for Team on annual plans.
This launch marks a significant step in OpenAI’s commitment to creating more agentic AI, where tools actively engage in problem-solving rather than merely responding to queries. As AI technology continues to advance, features like ChatGPT Agent are poised to redefine efficiency, particularly for tech-savvy users in Australia.
Stay tuned for more updates as this story develops, and for additional information, visit OpenAI’s official site.
