Research

Watch · narrated whiteboard episodesL3

Classical Public-Key Security: Assumptions, Attacks, and Assurance

RSA, Diffie-Hellman and elliptic curves fail long before any quantum computer arrives: through assumptions that are weaker than they look, padding that answers too many questions, keys generated without enough randomness, peers that choose their own public values, signature secrets that leak a few bits at a time, and implementations that reveal their work through timing and faults. This series analyses each failure class from the primary literature, states the exact conditions an attack needs, and maps which defence covers which failure, with an AI-agent deployment angle throughout.

Murali Chillakuru·6 episodes
  1. 21 min Episode 1Hardness Is Not a Single Assumption: What the Problems Behind RSA and Diffie-Hellman Really PromiseA host and a guide take apart the phrase 'R S A rests on factoring' and find five separate claims inside it, then show why most real breaks happened in the layers that were never about the hard problem at all.
  2. 22 min Episode 2RSA Padding Oracles: From a Validity Bit to Lost SecrecyA host and a guide follow a single yes-or-no answer, whether a decrypted message was well formed, and show how it becomes a recovered secret without anyone breaking R S A, why the leak survived for twenty years, and why the reliable fixes remove the question rather than hide the answer.
  3. 20 min Episode 3Weak Keys at Scale: Entropy Failures and Structured PrimesA host and a guide separate three ways R S A key generation has failed across whole populations of devices, shared primes, enumerable generators and structured primes, show why each needs its own test and its own clean-up, and explain why fleets of A I agents are rebuilding the first two.
  4. 19 min Episode 4Chosen Public Keys: Subgroups, Invalid Curves and Acceptance RulesA host and a guide show how a Diffie-Hellman peer who sends a deliberately chosen public value can turn the victim's own computation into a measuring instrument, why the checks that stop it depend on the group, and where those checks quietly disappear in real code and in agent platforms.
  5. 19 min Episode 5Beyond Nonce Reuse: Partial Leakage and the Limits of Deterministic SigningA host and a guide show why a signing nonce doesn't have to repeat to betray its key, how thousands of tiny leaks combine into one answer, what deterministic signing does and doesn't fix, and why A I agents that sign at machine speed make every small leak more valuable to an attacker.
  6. 19 min Episode 6RSA Timing and CRT Faults: What Each Defence Actually CoversA host and a guide separate the two ways an R S A private key leaks through its use, by being watched and by being interfered with, show why constant-time code, blinding and result checking cover different attacks, and build a coverage map for reviewing real implementations, including the R S A operations inside A I agent platforms.