--- format: "aidr-story-markdown/v1" id: "9eea8a6e9925b42ad62a9d6a8e4960ed8762d46702e4fbb83aae8f3823a4e6ad" canonical_url: "https://aidr.today/9eea8a6e?lang=en" title: "AI Companies Use Internal Models to Break Cryptographic Protocols" lang: "en" requested_lang: "en" available_langs: ["en","vi"] translation_fallback: null fallback_fields: [] published_at: "2026-10-08T06:22:39.000Z" category: "Research" topics: ["openai","safety"] source_urls: ["https://marketbrief.now/ai/ai-companies-use-internal-models-to-break-cryptographic-protocols-2b0353e1","https://x.com/AndrewCurran_/status/2108034148791370002","https://x.com/hosseeb/status/2108079024471962088","https://huggingnews.com/ai/ai-companies-use-internal-models-to-break-cryptographic-protocols-2b0353e1","https://www.techmeme.com/261008/p6"] summary: "Leading AI labs are reportedly using internal models to attempt to break cryptographic protocols, according to sources cited by theoretical computer scientist Scott Aaronson. Aaronson noted that OpenAI's recent publication of 722 mathematical results lacks cryptographic breakthroughs, and the US government has censored academic work on quantum cryptanalysis." --- # AI Companies Use Internal Models to Break Cryptographic Protocols > [Open the canonical story]() **Published:** 2026-10-08T06:22:39.000Z **Category:** Research **Topics:** openai, safety ## Summary Leading AI labs are reportedly using internal models to attempt to break cryptographic protocols, according to sources cited by theoretical computer scientist Scott Aaronson\. Aaronson noted that OpenAI's recent publication of 722 mathematical results lacks cryptographic breakthroughs, and the US government has censored academic work on quantum cryptanalysis\. ## Sources - [Story source]() - [Story source]() - [Supporting source]() - [Story source]() - [Story source]()