Technology covers a security-class shift: today’s web mostly authorizes under computational hardness assumptions that expire when machines, algorithms, or archives catch up; a different proof class aims for public transcripts that carry no secret-relevant information under stated axioms. The hub separates transport secrecy jobs (TLS, PQ hybrids) from authorization emptiness (Channel Zero, Rosario), so buyers and builders pick the right door before product SKUs.
For fifty years the internet has mostly authorized under computational hardness assumptions: certain math problems stay hard enough for transport and tokens. That bet powers TLS and the padlock feeling in the browser, but stolen credentials still authorize because possession travels with the secret. ENI6MA is not a harder padlock on that road. It is a different claim class: Channel Zero and Rosario-style proofs aim so observers can record the public ceremony and still learn nothing usable about the secret under named axioms.
Computational
Empty channel (ITS)
Two roads of security · different claim class
Read left-to-right: the computational road treats secrecy as a computational hardness assumption (TLS/PKI-style). The information-theoretic road treats authorization as an empty public channel: recording the transcript does not yield the secret. Takeaway: ENI6MA argues a different job, not a harder padlock on the same road.
Call the first road the computational bargain: security that assumes attackers are too slow to break the math. RSA, elliptic curves, and lattice schemes all live here. When machines get faster, algorithms improve, or an archive waits long enough, the bet can expire, and a stolen key still authorizes until someone revokes it.
Call the second information-theoretic authorization: security that does not rely on computers being too slow. It asks whether the public channel even carries secret-relevant information. If the transcript is empty of that information under the model, recording it is not learning the secret. Two roads: computational hardness assumptions versus an empty effective channel under the named model.
Computational security versus information-theoretic authorization for authorization transcripts
Dimension
Computational security
Information-theoretic authorization
Core bet
Hard problems stay hard (RSA, ECC, lattices…)
Public transcript reveals nothing usable about the secret
One-shot proofs across digital systems and paper. Authority expires with the request - under the reference architecture and claim model.
Passwords, tokens, and certificates keep working until someone rotates or revokes them - including an attacker who already has a copy. ENI6MA replaces that lasting authority with one-shot proofs: each approval authorizes one action once, across apps, agents, and paper. Captured proofs do not replay; authority ends with the request. Deploy as Public, Cloud, or Sovereign stacks under the reference architecture and model-scoped claims.
One-shot, request-bound proofs mean authority expires with the request. Replay has nowhere to go when burn-before-validate is enforced: the nonce is spent on the ledger before validation finishes. That is the product claim in one line: ENI6MA replaces reusable credentials with one-shot, request-bound proofs. Authority expires with the request under the reference architecture (claim: authority-expires-per-request).
Stored secret vs one-shot proof
Read left-to-right: the stored-secret column shows ambient credentials that keep working after copy; the one-shot column shows a request-bound envelope that burns after accept. Takeaway: stolen possession is not lasting authority when the proof dies with the moment.
Where it ships in a deployment is a Security and Products capability description. Security architecture covers gate order and coexistence; Technology states why the claim class differs.
Four doors into Technology
Nested navigation mirrors these four headers. Each door has a job statement and a primary CTA; start there, then follow the reading path if you want the full spine.
Fundamentals
Why proofs differ; the computational bet; the empty channel; recording without learning; and reading paths so you never start in the wrong chapter.
Papers, glossary and notation, model-scoped claims, and the engineering stack that connects theory to products, without burying the Fundamentals pages.
Foundry mints identity into circuits. Control registers, ledgers, and observes. Gate enforces at the boundary. Verify attacks your deployment the way an adversary would. That product plane is documented under Products and Security. Technology states why; Products and Security cover how it runs. Engineering stack detail for implementers lives at Engineering stack, not on this hub.
Computational security is usually stated as a hardness game: a bounded adversary fails except with negligible probability. Information-theoretic authorization for Channel Zero is stated as emptiness of mutual information on the public ceremony channel. Notation lives in the notation table.
Message perfect secrecy (Shannon spirit)
I(M;C)=0
I of M semicolon C equals zero
Read as: The ciphertext carries zero mutual information about the message.
: posterior belief does not change after observing ciphertext
Significance: It states the classical emptiness claim: ciphertext observation does not update beliefs about the message.
Computational contrast: The computational bargain replaces this emptiness requirement with a computational hardness assumption so strangers can talk at scale without OTP logistics.
: mutual information between hidden state and observation
∣\mid
: conditioning on the public challenge
=0= 0
: zero effective leakage under the stated model
Significance: Same Shannon spirit as OTP message secrecy; different job: authorization emptiness on the public ceremony channel.
Computational contrast: Hardness games ask whether a bounded adversary can forge or decrypt under a computational hardness assumption; this equation asks whether the transcript carries secret-relevant information at all.
