On the morning of July 14th, the list of Fields Medal winners, initially slated for announcement at the opening ceremony of the International Congress of Mathematicians (ICM 2026) in Philadelphia on July 23rd, was inadvertently disclosed due to a curl command generated by Codex.

This unintended revelation brought a delightful shock to China, as two Peking University alumni, Hong Wang and Yu Deng, were featured on the list. This marks the first time that Chinese mathematicians are poised to claim the highest accolade in mathematics in the same year.

Here's the backstory: the eventSnapshot interface, typically employed by the official website to load the schedule, inadvertently sent complete data for all meeting entries to the front end, including those marked as hidden. Among these were four Fields Medal winner lecture entries labeled as 'HIDDEN'.

A netizen utilized a curl command generated by Codex to fully capture this concealed entry, leading to the emergence of four names: Yu Deng, John Pardon, Jacob Tsimerman, and Hong Wang.

Furthermore, the corresponding award presenters for each winner, as well as the recipients of prizes such as the Shayan Oveis Gharan Prize (Abacus Prize), Graeme Segal Prize (Chern Prize), and Yurii Nesterov Prize (Gauss Prize), were also accessible from the same database.

The news rapidly circulated across Chinese platforms like Zhihu, with numerous users independently replicating the data scraping process.

The ICM official website subsequently experienced a brief outage, and upon restoration, all HIDDEN entries had been removed.

However, the data had already been widely disseminated. Notably, as a result of this news, the winning probabilities of the four individuals on the prediction market Polymarket soared to 98% within hours.

Hong Wang's most significant contribution was her collaboration with Joshua Zahl in solving the three-dimensional Kakeya conjecture in 2025. This long-standing problem, which had perplexed the mathematical community for over a century, posits that in three-dimensional space, a set containing unit line segments in all directions must have a Minkowski dimension and Hausdorff dimension equal to 3. Her proof, spanning 127 pages, introduced a novel methodology of 'granularity' analysis.

She is currently a professor at the Courant Institute of Mathematical Sciences at New York University and the first female permanent professor in the history of the Institut des Hautes Études Scientifiques (IHES) in France. From a small county in Guilin, Guangxi, to Peking University, École Polytechnique in Paris, and MIT, Hong Wang's journey exemplifies the internationalization of Chinese mathematical talent.

Yu Deng's breakthrough relates to an even older objective—Hilbert's sixth problem. Collaborating with Xiao Ma and Zaher Hani, he rigorously derived, for the first time, the complete logical chain from microscopic particle systems to the macroscopic Boltzmann equation. Lanford's 1975 proof was only valid for an extremely short duration and had remained stagnant for half a century thereafter. Deng and his colleagues systematically addressed the core obstacle of 'recollisions'.

The ICM 2026 official website had previously marked the sixth problem as 'partially solved' and highlighted the achievements of Yu Deng's team. He grew up in Shenzhen, won a gold medal at the IMO in 2006, and is currently a mathematics professor at the University of Chicago.

If the leaked list is officially confirmed on July 23rd, it will mark the first instance in the history of the Fields Medal where two Chinese mathematicians have been honored in the same year.

An interesting coincidence is that Lions and Yoccoz, who jointly won the award in 1994, had both entered the École Normale Supérieure in Paris in 1975, making them the only pair of undergraduate classmates to win the Fields Medal in its history. If Hong Wang and Yu Deng share the stage, they will become the second such pair.

Some netizens commented that the 2007 class of the School of Mathematical Sciences at Peking University has truly achieved legendary status.

Just three days prior to the leak, OpenAI announced that GPT-5.6 Sol Ultra had autonomously completed a full proof of the 'cycle double covering conjecture' in graph theory in under an hour. Sixty-four sub-agents collaborated in parallel, solving a problem that had eluded human mathematicians for half a century within that timeframe.

Consequently, some speculate that 2026 may be the final year when the Fields Medal is awarded exclusively to humans.

With nine days remaining until the opening of ICM 2026, the International Mathematical Union has not commented on the leak, and it remains uncertain whether the deleted HIDDEN entries represent the final list. However, the academic achievements of Hong Wang and Yu Deng do not require a leaked JSON file for validation; their solutions to the Kakeya conjecture and Hilbert's sixth problem are already indelibly inscribed in mathematical history.

A front-end bug sparked a sensation more impactful than any official press release.

Do you believe that AI independently completing mathematical proofs will aid humans in accelerating the resolution of difficult problems or undermine the credibility of the Fields Medal? Feel free to share your thoughts in the comments section.