Introduction
In a swift move reflecting its ongoing commitment to developing advanced products that meet growing public demands, OpenAI launched a comprehensive new update to its artificial intelligence model named “GPT-5.5 Instant” on Wednesday, June 24. This release marks the second behavioral update applied to the default model of the popular chatbot “ChatGPT” within a record timeframe of less than a single month, aiming to provide a smarter, more flexible, and more human-like user experience.
- Details of the latest behavioral update
- Upgrading the shopping and recommendation experience
- User interactions and reception of updates
- A rapid pace outpacing public opinion
- Frequently asked questions
Details of the latest behavioral update
The new update, which was included in the chatbot’s official release notes page, focuses primarily and intensively on enhancing the generative model’s ability to comprehend and understand the true intentions of users, equipping it with advanced skills to handle complex constraints and conditions presented in their daily queries. The initial launch of the GPT-5.5 Instant version on May 5 as a default model held promise by reducing digital hallucination rates by up to 52.5 percent in sensitive and high-risk fields, in addition to delivering answers that were nearly 30 percent shorter to become more focused and efficient. Less than four weeks later, on May 28, the company launched its first behavioral update, which made the model’s responses appear more conversational and natural, avoiding the excessive use of bulleted lists to make the text more fluid and interactive.
Upgrading the shopping and recommendation experience
The June 24 update goes much further, targeting the smart model’s persona and its practical usefulness in users’ daily lives. According to the company, this update became immediately available to paying subscribers in the Plus and Pro tiers, and is scheduled to be rolled out to free-tier users the following day. This update places utmost importance on improving recommendations related to daily shopping and local map queries, building on enhancements the company introduced to the commerce experience earlier this year. In March, the company upgraded the mechanism for retrieving and displaying product information using the “Agentic Commerce Protocol,” a smart system that enables detailed comparisons including prices, ratings, and specific product features. This update aims to make recommendation paths more cohesive and logical, especially when adding stringent constraints such as budget and geographic location.
User interactions and reception of updates
Michelle Pokras, head of the research team at the company, noted that the initial version of the model “relied too heavily and annoyingly on ready-made bullet points” and was in urgent need of further work and tuning in areas related to exaggeration, factual accuracy, and good multilingual performance. It is clear that the recent revision strongly continues this corrective path, focusing on building and designing a model that is less robotic and more capable of harmonizing with natural, unpretentious human conversation. However, this update has not always received widespread welcome from all users, as many pointed out in specialized groups that the model’s performance can appear erratic across these frequent revisions. Some users described the previous update as being heavily censored and less responsive to use cases requiring an understanding of subtle emotions.
A rapid pace outpacing public opinion
To date, OpenAI has not officially and publicly commented on the multiple user complaints, yet it continues to push forward with continuous updates, suggesting that the company’s internal metrics and analytical data are driving and controlling these changes at a pace much faster than public opinion and complaints might absorb. This rigorous approach ensures the continuous improvement of AI efficiency ahead of expectations.
Frequently asked questions
Question: What is the most prominent change in the GPT-5.5 Instant update?
Answer: The update focuses on making responses more natural and conversational, with a noticeable improvement in delivering shopping recommendations and managing user-posed constraints to save search time.
Question: When will free users receive this update?
Answer: The update was made available to paid subscribers first, and is scheduled to be provided to free-tier users the following day to ease technical pressure.
Question: What is the Agentic Commerce Protocol mentioned in the article?
Answer: It is a system developed by OpenAI to enable the model to accurately retrieve product information and conduct comparisons including prices and reviews to assist users in making purchases.
Question: Why did some users complain about the frequent updates?
Answer: Some complained about performance fluctuations and considered that the updates made the model stricter and less aware of subtle and innovative emotional details in certain use cases.
