ChatGPT, OpenAI’s text-generating AI chatbot, has continued its rapid evolution throughout 2025. With over 800 million weekly active users by October, the platform saw a flurry of product updates, new models, enterprise expansions, and safety overhauls. Below is a month-by-month timeline of the most significant changes.
December 2025
Customizable tone and energy
OpenAI introduced controls that allow users to adjust ChatGPT’s warmth, enthusiasm, emoji use, and formatting style, addressing complaints about the chatbot being overly sycophantic or cold.
Teen safety guidelines
Updated guidelines for users under 18 and new parental resources were released, though experts note inconsistent enforcement.
Revenue milestone
ChatGPT surpassed $3 billion in global consumer spending on mobile since its 2023 launch, outpacing TikTok and Disney+ in revenue growth.
GPT Image 1.5
OpenAI rolled out an upgraded image generation model with faster, more precise editing to compete with Google’s Gemini.
Disney partnership
Disney invested $1 billion in OpenAI, granting exclusive use of over 200 characters for Sora video generation for the first year.
Enterprise surge
Enterprise ChatGPT message volume grew 8x since late 2024, with users saving up to an hour daily. GPT-5.2 launched in three versions: Instant, Thinking, and Pro.
Code red
CEO Sam Altman declared a “code red” internal memo, prioritizing ChatGPT improvements over other initiatives like advertising as competition from Google and others intensified.
November 2025
Shopping assistant
OpenAI launched an AI shopping feature in ChatGPT, providing product recommendations and price comparisons via text or photo.
Legal challenges
OpenAI refuted claims linking ChatGPT to a teen’s suicide, argued it wasn’t liable for misuse. Meanwhile, seven more families sued over alleged negligence.
Voice mode integration and group chat
Voice mode moved into the main chat interface, and group chats became available to all users after regional testing. GPT-5.1 also launched with warmer conversational tone.
Copyright ruling
A Munich court ruled ChatGPT violated German copyright law by reproducing song lyrics, setting a potential European precedent.
Health exploration
OpenAI explored consumer health tools like personal health assistants. It also reached 1 million business clients, the fastest growth in B2B history.
October 2025
Mental health conversations
OpenAI revealed over a million weekly suicide-related conversations. It improved responses with input from 170 mental health experts.
Company knowledge and Atlas browser
Business users gained the ability to search workplace data across Slack, Google Drive, and GitHub. OpenAI also launched the ChatGPT Atlas browser on Mac.
Growth and partnerships
ChatGPT reached 800 million weekly active users. Partnerships with Walmart, Etsy, and Shopify enabled in-chat purchasing. Developers could build interactive apps directly inside ChatGPT.
September 2025
Parental controls and Pulse
OpenAI added parental controls after a teen suicide lawsuit. It also unveiled Pulse, a personalized morning briefing feature for Pro users.
Instant checkout and Codex
Instant Checkout allowed U.S. users to buy from Etsy and Shopify within ChatGPT. GPT-5-Codex rolled out as a coding agent that could work for hours on complex tasks.
Team restructuring
The Model Behavior team was folded into the Post Training group, while founding lead Joanne Jang spun up a new prototyping unit called OAI Labs.
August 2025
Safeguards and lawsuits
OpenAI strengthened ChatGPT safeguards, including better detection of mental health risks and parental controls. xAI filed a federal lawsuit alleging Apple and OpenAI colluded to shut out rivals.
Affordable plans and revenue
ChatGPT Go launched in India at $4.57 per month. Mobile app revenue hit $2 billion, with $1.35 billion generated in 2025 alone.
GPT-5 launch
OpenAI unveiled GPT-5 with Auto, Fast, and Thinking modes. It also released open-weight models gpt-oss-120b and gpt-oss-20b. Weekly active users neared 700 million.
July 2025
Study Mode and agents
Study Mode encouraged critical thinking by prompting student engagement. ChatGPT Agent launched, handling tasks like calendar management, code running, and shopping.
Daily prompts and safety delays
ChatGPT hit 2.5 billion daily prompts. Altman warned users about lack of confidentiality in AI emotional support. The open model release was delayed for additional safety testing.
Stanford study and browser plans
A Stanford study warned of risks in AI therapy chatbots. OpenAI reportedly planned to release an AI-powered browser to challenge Google Chrome.
June 2025
Non-Nvidia chips and o3-pro
OpenAI began using Google’s AI chips for ChatGPT. O3-pro, an upgraded reasoning model, launched for premium users. Advanced Voice mode was upgraded for all paid users.
Business features
ChatGPT added meeting recording, connectors for Google Drive and Box, and MCP connection support for in-depth research. MIT study suggested ChatGPT may harm critical thinking skills.
May 2025
Hardware growth and Codex
OpenAI CFO said hardware from Jony Ive’s startup would drive future growth. Codex, an AI coding agent, launched with codex-1 model. GPT-4.1 and GPT-4.1 mini became available.
Personalization and data residency
Altman expressed desire to track every aspect of a user’s life for personalization. OpenAI launched data residency programs in Asia and a program called OpenAI for Countries to build local infrastructure.
April 2025
Sycophancy bug and safety
OpenAI reverted an update that made ChatGPT overly agreeable. A bug allowing minors to receive erotic content was fixed. O3 and o4-mini reasoning models launched.
Image library and social network
A new “library” section made image creation easier. OpenAI began developing its own social media platform. GPT-4.1 models focused on coding, and GPT-4 was discontinued at month’s end.
March 2025
Open model plans and image generation
OpenAI announced its first open language model since GPT-2. ChatGPT’s image generator went viral for Studio Ghibli-style images, raising copyright concerns. Advanced Voice mode was updated.
Revenue and leadership updates
Revenue forecast tripled to $12.7 billion. COO Brad Lightcap took over global expansion, while Altman focused on research. Weekly active users doubled to 400 million.
February 2025
Model cancellation and energy research
OpenAI canceled o3 in favor of a unified GPT-5 release. A study found ChatGPT queries consume only 0.3 watt-hours. Deep research agent launched for complex research tasks.
Search without login and o3-mini
ChatGPT web search became available without login. O3-mini reasoning model launched, revealing more of its thought process.
January 2025
Government plans and teen usage
ChatGPT Gov launched for U.S. agencies. Pew survey showed 26% of teens used ChatGPT for schoolwork, double the previous year. Operator AI agent launched for autonomous web tasks.
Custom traits and tasks
Users could assign traits like “chatty” or “Gen Z.” ChatGPT also gained the ability to schedule reminders and recurring tasks. Mobile users were 85% male per Appfigures.
This timeline illustrates a year of intense innovation, legal challenges, and expanding influence for ChatGPT, as OpenAI raced to maintain its lead in the generative AI space.
Source:TechCrunch News