OpenAI to Develop a ‘Super Assistant’ Based on ChatGPT

OpenAI aims to transform ChatGPT into a ‘super assistant’ that ‘deeply understands the user and serves as an interface for the internet,’ according to court documents.

The materials were disclosed as part of the antitrust proceedings involving Google and the Department of Justice.

The company plans to implement this initiative in the first half of 2025. Representatives believe that models o2 and o3 have become sufficiently intelligent to perform agent tasks.

OpenAI describes the ‘super assistant’ as an intelligent entity that is personalized and universally accessible. It possesses the skills and deep knowledge to ‘perform daily tedious tasks’: answering questions, finding housing, connecting with a lawyer, booking a gym session, planning vacations, purchasing gifts, managing calendars, and sending emails.

The AI startup led by Sam Altman views other chatbots as competitors: Claude, Gemini, Copilot, Meta AI. The company considers itself a leader and acknowledges the need for further development.

‘We have everything needed to win: one of the fastest-growing products of all time, a category-defining brand, leadership in research (reasoning, multimodality), computing, a world-class research team, and a growing number of effective people motivated to work,’ states a document dated late 2024.

The company emphasized the need for healthy competition, which will benefit everyone. Users should have the right to choose their AI assistant, and Apple, Google, and Microsoft should not promote their own solutions ‘without offering fair alternatives.’

In May 2025, it was reported that Altman intended to create 100 million ‘AI companions’—devices that would become part of users’ daily lives.