PulseAugur
EN
LIVE 18:52:41

Anthropic's Claude model finds cryptographic weaknesses, launches new benchmark · 9 sources tracked

By PulseAugur Editorial · [9 sources] ·

Anthropic's Claude Mythos Preview model has demonstrated advanced capabilities in cryptography research by autonomously discovering weaknesses in both a digital signature scheme called HAWK and a reduced version of the Advanced Encryption Standard (AES). While these findings do not currently impact deployed systems, they highlight the potential of frontier AI models to perform expert-level cryptographic analysis, which could have significant defensive applications. Anthropic collaborated with academics from ETH Zurich, Tel Aviv University, and the University of Haifa to develop CryptanalysisBench, a new benchmark for evaluating LLMs' cryptanalysis abilities. AI

IMPACT Demonstrates frontier AI models' potential for expert-level cryptography research, with implications for defensive cybersecurity applications.

RANK_REASON Research paper and benchmark release by an AI lab.

Read on X — Anthropic →

AI-generated summary · Google Gemini · from 9 sources. How we write summaries →

Anthropic's Claude model finds cryptographic weaknesses, launches new benchmark · 9 sources tracked

COVERAGE [9]

  1. X — Anthropic TIER_1 English(EN) · AnthropicAI ·

    We also worked with academics at ETH Zurich, Tel Aviv University, and the University of Haifa to build CryptanalysisBench, a benchmark for studying LLMs’ crypta

    We also worked with academics at ETH Zurich, Tel Aviv University, and the University of Haifa to build CryptanalysisBench, a benchmark for studying LLMs’ cryptanalysis abilities. https://t.co/OEYhGjH5Lj

  2. X — Anthropic TIER_1 English(EN) · AnthropicAI ·

    Full technical details of both attacks are provided in our new papers:

    Full technical details of both attacks are provided in our new papers: On HAWK: https://t.co/JtLzetLOtU On AES: https://t.co/7pozPjc4o8 And the associated model chain-of-thought for AES: https://t.co/Cg1tfTJhvi

  3. X — Anthropic TIER_1 English(EN) · AnthropicAI ·

    Still, both results show that frontier AI models are capable of doing expert-level cryptography research. This has important defensive applications—testing the

    Still, both results show that frontier AI models are capable of doing expert-level cryptography research. This has important defensive applications—testing the algorithms that keep our online activity secure, and ultimately helping to make digital systems safer.

  4. X — Anthropic TIER_1 English(EN) · AnthropicAI ·

    These are substantial research advances, but they don’t have a practical impact on today’s systems. HAWK is a proposed scheme that hasn’t been deployed anywhere

    These are substantial research advances, but they don’t have a practical impact on today’s systems. HAWK is a proposed scheme that hasn’t been deployed anywhere, and the AES attack we discovered was on a weaker version and does not break the full cipher.

  5. X — Anthropic TIER_1 English(EN) · AnthropicAI ·

    Mythos Preview did most of this work autonomously, with occasional human guidance. Each of the two results cost roughly $100,000 in API usage.

    Mythos Preview did most of this work autonomously, with occasional human guidance. Each of the two results cost roughly $100,000 in API usage. We disclosed the findings in advance to the algorithms’ authors, as well as to US government and industry partners.

  6. X — Anthropic TIER_1 English(EN) · AnthropicAI ·

    The symmetric cipher is a reduced version of the Advanced Encryption Standard (AES)—which has received decades of scrutiny (more than almost any other encryptio

    The symmetric cipher is a reduced version of the Advanced Encryption Standard (AES)—which has received decades of scrutiny (more than almost any other encryption algorithm). In a week, Mythos Preview found a way to speed up an attack on this version of AES by 200-800×.

  7. X — Anthropic TIER_1 English(EN) · AnthropicAI ·

    The digital signature scheme is HAWK, which is designed to be robust even against hypothetical quantum computers.

    The digital signature scheme is HAWK, which is designed to be robust even against hypothetical quantum computers. HAWK has survived two years of expert review, but in 60 hours Mythos Preview found a previously-unknown attack that reduced the scheme’s key strength by half.

  8. X — Anthropic TIER_1 English(EN) · AnthropicAI ·

    Claude discovered weaknesses in a highly-secure digital signature scheme (used to verify identity digitally) and a well-known symmetric cipher (used to encrypt

    Claude discovered weaknesses in a highly-secure digital signature scheme (used to verify identity digitally) and a well-known symmetric cipher (used to encrypt data).

  9. X — Anthropic TIER_1 English(EN) · AnthropicAI ·

    New Anthropic research: Discovering cryptographic weaknesses with Claude.

    New Anthropic research: Discovering cryptographic weaknesses with Claude. Claude Mythos Preview has helped our researchers find weaknesses in cryptographic algorithms—the mathematical methods that are used to keep data private. Read more: https://t.co/TYKLjb3Q7V

RELATED ENTITIES

RELATED TOPICS