How OpenAI’s Latest Model Will Impact ChatGPT
OpenAI is having an incredible day today. First, the company announced a $1 billion investment from Disney, along with a licensing agreement that will allow Sora users to create videos featuring characters like Mickey Mouse, Luke Skywalker, and Simba. Shortly after, OpenAI unveiled its latest major language model: GPT-5.2 .
OpenAI claims its new GPT model is particularly useful for “professional intellectual capital work.” It touts GPT-5.2 as superior to previous models in spreadsheet creation, presentation preparation, coding, image analysis, and multi-stage project management. To build this model, the company also collected data from tech companies: Notion, Box, Shopify, Harvey, and Zoom credit GPT-5.2 with “cutting-edge long-term reasoning,” while Databricks, Hex, and Triple Whale believe GPT-5.2 is “exceptional” in both agent-based data analysis and document analysis tasks.
However, most OpenAI users are non-professionals. Most users who will interact with GPT-5.2 are using ChatGPT, many of them for free. What can these users expect when OpenAI updates the free version of ChatGPT with these new models?
How GPT-5.2 Works in ChatGPT
OpenAI claims that GPT-5.2 will improve ChatGPT’s everyday functionality. The new model supposedly makes the chatbot more structured, reliable, and “pleasant to interact with,” though I’ve never seen evidence of this last part.
GPT-5.2 will impact ChatGPT’s performance differently depending on which of the three models you use. According to OpenAI, GPT-5.2 Instant is designed for “everyday work and learning.” It appears to be better suited for questions related to topic-specific information retrieval, questions and step-by-step instructions, technical writing, and translation—perhaps ChatGPT will make you give up your Duolingo obsession.
However, the GPT-5.2 Thinking model is claimed to be designed for “deeper work.” OpenAI wants you to use this model for programming, summarizing long documents, answering file requests sent to ChatGPT, solving math and logic problems, and making decisions. Finally, there’s GPT-5.2 Pro, OpenAI’s “smartest and most robust option” for tackling the most complex tasks. The company claims that 5.2 Pro produces fewer errors and demonstrates improved performance compared to previous models.
GPT-5.2 Security Policy
OpenAI claims that this latest update improves the models’ response to alarming queries, such as signs of suicidal tendencies, self-harm, or emotional dependence on AI. As a result, the company says, the GPT-5.2 Instant and Thinking models exhibit “fewer unwanted responses” compared to GPT-5.1 Instant and Thinking. Furthermore, the company is working on an “age prediction model” that will automatically set content restrictions for users it deems to be under 18.
These safety improvements are important—even critical—as we begin to understand the relationship between chatbots and mental health. The company acknowledged its inability to ” recognize signs of delusions ” as users turned to the tool for emotional support. In some cases, ChatGPT fueled delusional thinking, encouraging dangerous beliefs. Some families have even sued companies like OpenAI, claiming their chatbots aided or abetted victims to commit suicide.
While it’s certainly good to actively acknowledge improvements in user safety, I think companies like OpenAI still have a long way to go and a lot to do.
OpenAI announced that GPT-5.2 Instant, Thinking, and Pro will be available for paid plans today, Thursday, December 11th. Developers will also be able to access the new models via the API starting today.
Note: In April, Lifehacker’s parent company, Ziff Davis, filed a lawsuit against OpenAI, alleging that it infringed Ziff Davis’s copyright in the training and operation of its artificial intelligence systems.