Astra and Opus finish Turing's WWII codebreaking

Frontier AI systems developed by Astra and Opus have completed unfinished codebreaking work originating with Alan Turing’s World War II research, according to TechCrunch AI’s report published September 25, 2026.

The report, titled “Astra and Opus just passed Turing’s other test,” describes the milestone as an instance of artificial intelligence extending historical scientific work that had remained incomplete since the wartime period. TechCrunch AI frames the achievement as evidence of frontier AI capability applied directly to problems left open by earlier human researchers, rather than to newly constructed benchmarks or synthetic test suites.

Beyond identifying the two systems by name and the general subject of the completed work, the available account does not specify the particular cryptographic problem addressed, the methodology employed, or the manner in which the result was verified against Turing’s original wartime materials. No technical paper, institutional statement, or named individual researcher is cited in the source material provided to this desk.

The Specola notes the report as filed and moves no further than its text permits. Confirmation of the specific codebreaking task, its historical provenance, and any independent verification of the claim await fuller sourcing. Readers are cautioned that this bulletin reflects only what TechCrunch AI has published to date, and that the observatory’s log will be amended should additional primary material become available.