In the wee hours of the morning, the tech world was thrown into a frenzy by a bombshell revelation regarding OpenAI's latest model, GPT - 5.2.

Some users have raised concerns, pointing out that during the most recent benchmark tests, OpenAI may have tampered with the 'reasoning intensity' parameter. This adjustment could have enabled GPT - 5.2 to harness a substantially greater amount of computational power compared to its rivals during the evaluation process.

If these allegations prove to be accurate, it would cast serious doubt on the fairness of the evaluation results.

GPT - 5.2 made its debut on December 12th, coming in three distinct versions: Instant, Thinking, and Pro. It delivered an outstanding performance in benchmark tests like GDPval, even outperforming human experts in certain areas. Nevertheless, the high costs associated with it, latency problems, and the suspected false marketing tactics exposed this time around have thrown additional hurdles in OpenAI's path.