How OpenAI’s Latest Model Will Impact ChatGPT

OpenAI is having a hell of a day. First, the company announced a billion-dollar equity investment from Disney, as well as a licensing deal that will allow Sora users to generate videos with characters like Mickey Mouse, Luke Skywalker, and Simba. Shortly after, OpenAI unveiled its latest major language model: GPT-5.2.

OpenAI claims that this new GPT model is particularly useful for “professional knowledge work.” The company announces that GPT-5.2 is better than previous models for creating spreadsheets, preparing presentations, writing code, analyzing images, and working on multi-step projects. For this model, the company also gathered information from tech companies: Notion, Box Shopify, Harvey, and Zoom are all speculated to find GPT-5.2 to have “cutting-edge long-term reasoning,” while Databricks, Hex, and Triple Whale believe GPT-5.2 to be “outstanding” with both agent data science and document analysis tasks.

But most OpenAI users are not professionals. Most users who will interact with GPT-5.2 use ChatGPT, and many of them for free. What can these users expect when OpenAI upgrades the free version of ChatGPT with these new models?

How GPT-5.2 works in ChatGPT

OpenAI claims that GPT-5.2 will improve the “everyday” functionality of ChatGPT. The new model is supposed to make the chatbot more structured, more reliable and “fun to chat with”, although I’ve never found the last part to be necessarily true.

GPT-5.2 will impact the ChatGPT experience differently depending on which of the three models you use. According to OpenAI, GPT-5.2 Instant is intended for “everyday work and learning.” This is apparently best for questions seeking information on certain topics, practice questions and walkthroughs, technical writing, and translations. Maybe ChatGPT will make you give up your Duolingo obsession.

However, GPT-5.2 reflection is supposedly designed for “deeper work.” OpenAI wants you to use this model for coding, summarizing long documents, answering queries on the files you send to ChatGPT, solving math and logic problems, and making decisions. Finally, there is GPT-5.2 Pro, OpenAI’s “smartest and most reliable option” for the most complex questions. The company claims that 5.2 Pro produces fewer errors and better performance compared to previous models.

GPT-5.2 Security Policy

OpenAI claims that this latest update improves how models respond to distressing prompts, such as those showing signs of suicide, self-harm, or emotional dependence on the AI. As such, the company claims that this model has “less unwanted responses” in GPT-5.2 Instant and Thinking compared to GPT-5.1 Instant and Thinking. Additionally, the company is working on an “age prediction model,” which will automatically impose content restrictions on users who, according to the model, are under 18 years old.

These security improvements are important, even critical, as we begin to understand the correlations between chatbots and mental health. The company admitted its failure to “recognize the signs of delusion,” as users turned to the tool for emotional support. In some cases, ChatGPT fueled delusional thoughts, encouraging people’s dangerous beliefs. Some families have even sued companies like OpenAI over allegations that their chatbots aided or encouraged victims to commit suicide.

Actively recognizing improvements in user security is undoubtedly a good thing, but I think companies like OpenAI still have a long way to go and a long way to go.

OpenAI says GPT-5.2 Instant, Thinking, and Pro will all roll out today, Thursday, December 11, to paid plans. Developers can also access the new models in the API today.

Disclosure: Lifehacker’s parent company, Ziff Davis, filed a lawsuit against OpenAI in April, alleging that it violated Ziff Davis’ copyrights in the training and operation of its AI systems.