Canadian mathematician Jacob Tsimerman, who recently received the Fields Medal, announced the creation of Mathematical A.I. Safety Institute (MAISI). This was reported by The Decoder in an article dated 11 September 2026 year.
The name of the initiative signals an attempt to verify AI safety using approaches comparable to proving the resilience of cryptographic codes. However, available information is presented only as metadata and a brief synopsis, so program details are not yet confirmed.
editorial commentary
Why it matters
The probable value of the initiative is to strengthen formal methods in AI safety research. The next verifiable signal will be the publication of a description of the methodology, team composition, or initial results. Significant uncertainty remains: current data do not confirm either a specific program or practical effectiveness of the approach.