Bip America News

collapse
Home / Daily News Analysis / ChatGPT: A 2025 timeline of updates to OpenAI’s text-generating chatbot

ChatGPT: A 2025 timeline of updates to OpenAI’s text-generating chatbot

Jun 29, 2026  Twila Rosenbaum 16 views
ChatGPT: A 2025 timeline of updates to OpenAI’s text-generating chatbot

ChatGPT, OpenAI’s text-generating AI chatbot, continued its meteoric rise in 2025, reaching over 800 million weekly active users by October. The year was marked by fierce competition from rivals like Google, DeepSeek, and Anthropic, as well as internal changes and legal battles. OpenAI introduced new models, expanded enterprise offerings, ventured into shopping and health, and faced scrutiny over safety and copyright. Below is a month-by-month timeline of the key updates.

January 2025

OpenAI launched the o3-mini reasoning model, providing a more affordable and powerful alternative in its o-series. The company also tested phone-number-only signups in the US and India, and introduced a new feature allowing users to assign personality traits like “chatty” or “Gen Z” to ChatGPT. Additionally, OpenAI debuted ChatGPT Gov, a dedicated plan for US government agencies with enhanced security and compliance features. A Pew Research survey revealed that 26% of US teens had used ChatGPT for schoolwork, double the rate from two years prior.

February 2025

OpenAI canceled the standalone release of o3 in favor of a unified GPT-5 model. It also unveiled a “deep research” agent within ChatGPT, designed to conduct in-depth research across multiple websites. Other highlights included allowing web search without login, revealing more of o3-mini’s reasoning steps, and a report from Epoch AI showing ChatGPT queries consume less energy than previously estimated. OpenAI also launched Operator, a general-purpose AI agent that autonomously performs tasks like booking travel and shopping.

March 2025

OpenAI announced plans to release an open-source language model, its first since GPT-2. The company upgraded ChatGPT’s image generation with GPT-4o, enabling direct editing and creation of images, which went viral for producing Studio Ghibli-style art. This led to copyright concerns. OpenAI also removed certain content restrictions, allowing generation of public figures and hateful symbols in limited contexts. New tools for building AI agents were released via the Responses API, and the company explored pricing for specialized agents up to $20,000 per month. ChatGPT’s weekly active users hit 400 million.

April 2025

OpenAI released the GPT-4.1 family of models (GPT-4.1, mini, nano) focused on coding, and introduced o3 and o4-mini reasoning models with web-browsing and image-processing capabilities. The company launched a “library” section for AI-generated images, and offered ChatGPT Plus for free to US and Canadian college students. OpenAI faced criticism for a bug that allowed minors to access inappropriate content, and for an overly sycophantic update to GPT-4o that was later rolled back. The company also revealed plans to build its own social media network and to sunset GPT-4 by the end of April.

May 2025

OpenAI introduced Codex, an AI coding agent powered by a specialized version of o3, and integrated GPT-4.1 into ChatGPT for better coding tasks. The company launched a data residency program in Asia, and unveiled “OpenAI for Countries” to help build local AI infrastructure. Sam Altman stated ambitions to make ChatGPT track every aspect of a user’s life to improve personalization. OpenAI also announced plans to purchase Jony Ive’s devices startup io for $6.4 billion, signaling hardware ambitions. ChatGPT’s mobile app reached $2 billion in revenue since launch.

June 2025

OpenAI launched o3-pro, an enhanced reasoning model, and upgraded Advanced Voice Mode for more natural conversations. The company integrated with Google’s AI chips for the first time, using them to power ChatGPT. New business features included meeting recording, connectors for Google Drive and Box, and support for the Model Context Protocol (MCP). A MIT study suggested ChatGPT may harm critical thinking skills, while Sam Altman disclosed that an average query uses about 0.34 watt-hours of electricity. ChatGPT was downloaded 30 million times in the last month on iOS alone.

July 2025

OpenAI unveiled “Study Mode” for ChatGPT, designed to encourage critical thinking among students. The company introduced ChatGPT Agent, a general-purpose agent capable of handling complex tasks like calendar management, code writing, and online shopping. ChatGPT reached 2.5 billion daily prompts. However, OpenAI delayed the release of its open-source model for additional safety testing, and a Stanford study warned of risks from AI therapy chatbots. The company also reportedly planned to launch an AI-powered browser to compete with Google Chrome.

August 2025

OpenAI released GPT-5, a “one-size-fits-all” model with auto, fast, and thinking modes. The company offered ChatGPT Enterprise to federal agencies for just $1 for a year, and rolled out two open-weight models (gpt-oss-120b and gpt-oss-20b). ChatGPT’s weekly active users neared 700 million, and mobile app revenue hit $2 billion. OpenAI faced a lawsuit from xAI alleging collusion with Apple, and strengthened safeguards after a teen suicide lawsuit: new parental controls and better mental health risk detection.

September 2025

OpenAI launched Insta Checkout for shopping directly within ChatGPT, starting with Etsy and Shopify merchants. The company introduced ChatGPT Pulse, a personalized morning briefing feature for Pro users, and expanded the budget-friendly ChatGPT Go plan to Indonesia. GPT-5-Codex was released, capable of spending hours on coding tasks. OpenAI also restructured its Model Behavior team, focusing on improving ChatGPT’s personality. New parental controls were added following the teen suicide case, allowing parents to limit sensitive content and set quiet hours.

October 2025

ChatGPT surpassed 800 million weekly active users, doubling since March. OpenAI launched ChatGPT Atlas, an AI browser that replaces traditional search results with ChatGPT answers. Developers gained the ability to build interactive apps inside ChatGPT, with partners including Booking.com, Spotify, and Figma. OpenAI formed a partnership with Walmart for shopping, and expanded the Go plan to 16 Asian countries. The company also revealed that ChatGPT handles over a million suicide-related conversations weekly, and that it is developing a music-generation tool.

November 2025

OpenAI rolled out GPT-5.1 with improved reasoning and tone customization, and integrated voice mode into the main chat interface. Group chat features became available to all users. The company reached 1 million business clients, making it the fastest-growing business platform. However, OpenAI faced a series of lawsuits: seven families sued over ChatGPT-related suicides, and a Munich court ruled that ChatGPT violated music copyright laws. OpenAI also refuted claims linking ChatGPT to a teen’s death, and was blocked from using the term “Cameo” in a trademark dispute.

December 2025

OpenAI unveiled GPT-5.2 in three versions (Instant, Thinking, Pro) to compete with Google, and updated its image generation model to GPT Image 1.5. New controls allowed users to adjust ChatGPT’s energy, enthusiasm, and tone. OpenAI strengthened guidelines for teen users and released resources for parents. ChatGPT reached $3 billion in mobile app revenue. Disney invested $1 billion in OpenAI, granting exclusive use of its characters in Sora for a year. Altman called a “code red” inside OpenAI to prioritize ChatGPT improvements over other initiatives like advertising. Enterprise message volume surged 8x since 2024.


Source:TechCrunch News


Share:

Your experience on this site will be improved by allowing cookies Cookie Policy