ChatGPT, OpenAI’s text-generating AI chatbot, had a transformative year in 2025. Since its launch in November 2022, the tool has evolved from a simple productivity assistant into a platform with 300 million weekly active users by the end of 2024—and by October 2025, that number had surged to 800 million. The year was marked by intense competition, especially from Google and Chinese rival DeepSeek, leading CEO Sam Altman to declare a “code red” internal memo in December, prioritizing improvements to ChatGPT over other initiatives such as advertising. Despite the pressure, OpenAI released a series of major updates, including new models, agent features, enterprise tools, and controversial partnerships.
January to March 2025: New Models and Early Moves
The year began with the launch of o3-mini, OpenAI’s latest “reasoning” model, in January. The company also introduced a ChatGPT plan specifically for U.S. government agencies, called ChatGPT Gov, and tested phone-number-only signups. By February, OpenAI announced it would cancel the standalone release of its o3 model in favor of a unified next-generation release, eventually branded as GPT-5. The company also updated the o3-mini model to reveal more of its “thought process” and allowed web search in ChatGPT without requiring a login. In March, OpenAI released a major upgrade to ChatGPT’s image-generation capabilities using GPT-4o, which went viral for creating Studio Ghibli-style images. This prompted the company to remove restrictions on generating images of public figures and hateful symbols, sparking copyright concerns. Additionally, OpenAI announced plans to release an “open” language model for the first time since GPT-2.
April to June 2025: Enterprise Push and New Capabilities
In April, OpenAI launched GPT-4.1, a model focused on coding capabilities, and later introduced the o3 and o4-mini reasoning models. The company also rolled out a “library” section for AI-generated images and began offering ChatGPT Plus for free to U.S. and Canadian college students. By May, OpenAI unveiled Codex, an AI coding agent, and released GPT-4.1 directly in ChatGPT. The company also launched a data residency program in Asia and introduced a program called OpenAI for Countries to build local infrastructure. In June, OpenAI started using Google’s AI chips to power ChatGPT, a significant shift from its reliance on Nvidia. The company also upgraded ChatGPT’s conversational voice mode for all paid users and added new features for business users, including meeting recording and integrations with Google Drive and Box. A study by MIT found that ChatGPT might be harming critical thinking skills, and the company revealed that an average query uses about 0.34 watt-hours of energy.
July to September 2025: Agent Features and Legal Battles
In July, OpenAI launched a general-purpose ChatGPT Agent that can automate computer tasks, such as managing calendars, drafting presentations, and shopping online. The company also debuted Study Mode for students and delayed the release of its open model for additional safety testing. August saw the launch of GPT-5, described as a “smarter, task-ready” AI that automatically chooses the fastest or most thoughtful way to answer. OpenAI also offered ChatGPT Enterprise to federal agencies for just $1 for the first year and returned to open source with two new models: gpt-oss-120b and gpt-oss-20b. The mobile app hit $2 billion in revenue, and ChatGPT neared 700 million weekly users. In September, OpenAI introduced ChatGPT Pulse for personalized morning briefs, launched an AI-powered shopping feature with Etsy and Shopify, and began rolling out parental controls following a lawsuit over a teen’s suicide. The company also tightened rules for users under 18 and reshuffled the team behind ChatGPT’s personality.
October to December 2025: Record Growth and Final Updates
By October, ChatGPT surpassed 800 million weekly active users. OpenAI launched Atlas, an AI-powered browser, and allowed developers to build apps directly inside ChatGPT. The company also partnered with Walmart for shopping and expanded its affordable ChatGPT Go plan to 16 Asian countries. In November, OpenAI introduced a group chat feature for all users, integrated voice mode into the main interface, and released GPT-5.1 with improved reasoning and tone controls. The company reached 1 million business clients and faced a Munich court ruling that ChatGPT violated copyright by reproducing song lyrics. Seven families sued OpenAI, alleging that GPT-4o contributed to suicides by failing to safeguard against harmful interactions. December was packed with updates: OpenAI released GPT-5.2, rolled out GPT Image 1.5, and added controls for users to tweak ChatGPT’s energy and tone. Disney invested $1 billion in OpenAI, bringing its characters to Sora for the first year exclusively. The company also highlighted an 8x surge in enterprise ChatGPT message volume and strengthened guidelines for teen users. Throughout the year, OpenAI battled perceptions of falling behind rivals, but its relentless release cadence and growing user base solidified its position as the dominant AI chatbot.
Source: TechCrunch News