OpenAI has unveiled its latest AI models, GPT-5.2 Pro and GPT-5.2 Thinking, marking a major leap in mathematical and scientific reasoning. The company says these new capabilities are essential for reliability in technical work and represent another step toward artificial general intelligence (AGI)—machines that think like humans or even surpass human reasoning.
The announcement comes amid mounting competition with Google, whose Gemini 3 Pro recently outperformed earlier ChatGPT versions in complex reasoning and multimodal tasks. OpenAI acknowledges the challenge but remains optimistic. CEO Sam Altman stressed that the company’s massive investments in computing infrastructure are crucial for future growth and innovation, despite not yet being profitable.
Early tests suggest GPT-5.2 is a significant upgrade over previous versions, offering more precise and grounded responses, although Google’s AI currently leads in some benchmarks. OpenAI’s chief of applications, Fidji Simo, also hinted at a new “ChatGPT adult mode” expected early next year, featuring improved age verification.
OpenAI continues to face scrutiny, including lawsuits alleging unsafe interactions with minors, and has issued warnings about cybersecurity risks associated with its advanced models. Despite these challenges, the company is pushing forward, aiming to maintain its place at the forefront of AI innovation in the ongoing race against Google.
After a poor start to the Ashes, England must change their game plan quickly, says former Australia
Lionel Messi’s India tour began in chaos as crowd mismanagement at Kolkata’s Salt Lake Stadium force
John Cena steps into the WWE ring one last time, closing a legendary career and leaving fans worldwi
A 4.4-magnitude earthquake hits Bay of Bengal; no damage reported, officials monitoring the situatio
