Report: OpenAI Could Launch ChatGPT 5 in the Coming Months

OpenAI is scheduled to release a new ChatGPT model in the coming months. The version, called GPT-5, could be released by the summer, according to a Business Insider report. Two sources close to Sam Altman, head of the AI Company, told Business Insider that some companies received demos of the advanced and improved ChatGPT model.

One of the CEOs who tested the GPT-5 gave his approval to the model, saying, “It’s really good, like materially better.” The CEO mentioned that “OpenAI presented a new model for their company with unique use cases and data.”

OpenAI is still working on GPT-5. Once OpenAI’s in-house team has completed a new multimodal broad language model, they will run it through the red conglomerate. Red teaming is a cybersecurity review process, during which a group of outsiders challenge the program and find problems or weak points in it that may have been overlooked by the authors.

A source told Business Insider that since there is no fixed timetable for conducting security testing; For this reason, the launch of ChatGPT-5 may be delayed, especially if the reading teams find flaws in the system.

ChatGPT’s main source of revenue comes from companies that pay OpenAI to provide them with an improved or customized version of ChatGPT. With CPT-5, the OpenAI team hopes to impress their potential customers and the public. ChatGPT was launched on November 30, 2022. Since its publication, it has grown in many directions and influenced many industries, from education to customer service.

Although OpenAI has not announced GPT-5, Sam Altman introduced Sora, an AI-powered tool that can generate videos based on text prompts. These videos can be one minute long. On February 15, Altman asked his X followers (formerly Twitter) to give him video prompts for Sora. He then shared the videos Sora made on his X account.

Despite its impressive results, the OpenAI team admits that Sora still has many weaknesses. The research organization wrote in a blog post that Sora “may struggle with accurately simulating the physics of a complex scene and may not understand certain cases of cause and effect.”

It added that the model might mix up “prompt state information, for example mixing left and right.” OpenAI also emphasized that it will ensure that Sora rejects prompts that violate their terms of use regarding hateful images, sexual content, or IP theft.