N [63) Key Event Receipt Infrastructure An Identity System Security Overlay for the Internet A Trust Spanning Layer for the Internet Samuel M. Smith Ph.D. https://keri.foundation https://keri.one/resources/
sam@keri.one
Documentation: https:ffkeri.foundationjresources| https:flkeri.onefkert resources/ KERI Suite Community: (meetings, open source code Apachez) https:figithub.com/WebOfTrust hetpsigithub.comWebOfTrust/keri TolP: KSWG, OWF License, within Linux Foundation, Decentralized Trust (meetings, specifications) ‘https://trustoverip.org/ hitps:{lif-toip.atlassian.netjwikilspaces/HOME/pages(56819755/KERI+Suites Working+Group Protocols and Specifications: CESR: https/github.comjtrustoveriptswg cesr specification KERI: https:/jgithub.comtrustoverip/tsw kerfspecifiation?tab=readme-ov-file ACDC: https:igithub.comtrustoveripftswe-acdc-specification TSP: https:/github.com/trustoveripjtswg-tsp-specification VLE: IS0 VVP: https;flwww.ietf.orglarchivelid/draft-hardman-verifiable-voice-protocokoo.html RAET: https://github.com/RaetProtocolfract (2013)
Adoptions GLEIF (VLEI): Organizational Identity (FSB Finance) & RERI https://www.gleif.org/en/lei-solutions/gleifs-digital-strategy-for-the-lei/ Cliinux introducing-the-verifiable-lei-vlei ngxs‘; EBA P3DH: Banking e https://www.eba.europa.eu/risk-and-data-analysis/pillar-3-data-hub @ DirectTrust: HISP Certification and vLEI ehatss https://directtrust.org/ ey healthKERI: SASE for Health Care Data @ oirectrrust https://healthkeri.com/ health<=RlI Provenant (GSMA): Telecom spam VVP and vLEI G PROVENANT https://provenant.net/ 2% Microsoft Cardano Foundation (Veridian): (Web3 Identity Layer) GSMA https://cardanofoundation.org/veridian i Pointanon Utah State: SEDI State Endorsed Digital Identity S.B. 260 https://le.utah.gov/Session/2025/bills/introduced/SBo260.pdf @ "!
User Public/Private Key Management:
Then (< 2015) and Now @ Then: impossibility of user-managed private keys Result: federated identity: IDPs with PKI-DNS/CA, OAuth N\ 4 Now: mobile devices/HSMs with MFA, secure boot/ enclaves, password management apps Result: practical user-managed private keys /\ Then: good enough security from federated identity f/\v% Result: session security, perimeter security, weak " y auth-crypt for AuthN Now: identity based on user-managed private key(s) Result: KERI
Security First, Always [NOIMORE| HALFIMEASURES] Minimally Sufficient Means Adoptable White Magic Crypto
First Principles Approach to the Universal Secure Attribution Problem Establish authorship of data, documents, credentials, entitlements, ... Authentic data provenance by anyone to anyone from anyone, i.e. universal Cryptographically verifiable, non-repudiable secure attribution of any communication to its source Solve authentic data provenance via universal secure attribution | controlling key(s) | | — secure identifier ‘«— authorship —«> — attribution
Identity (-ifier) System Security Overlay Establish authenticity (securely attribute source) of message payload. Authenticatable Message Verifiable Payload i identifier data signature j B S [ identier |+~ identiy system mapping |« ~| key-pair | The overlay’s security is contingent on the mapping’s security.
Private Key Management | controling key(s) | R identif o autorship = date [ identier | ey ata The continued use of private keys exposes them to side-channel attacks. This exposure weakens private keys over time (as the likelihood of a successful attack rises). Thus, from time to time, private keys must be revoked and replaced, i.e., rotated Conventional PKI must re-establish the root-of-trust with each rotation Thereby making rotation highly vulnerable to attack Compromised private keys result in loss of control over the identifier This breaks the chain-of-trust-of-control over the identifier Key compromise recovery must detect and regain control over the identifier Key rotation with key compromise recovery is the hard problem of private key management KERI solves this hard problem
Identifier Theory controlling key(s) | L sam identiier |- atanty < dola Short-term key-pair with public key as identifier = ephemeral identifier Long-term key-pair with public key as identifier = less ephemeral identifier Cryptographically derived identifier controlled by rotatable key-pair(s) = persistent (unbounded-term) identifier
Trust Basis of a Trust Domain Atrust basis binds controllers, identifiers, and key-pairs. A trust domain is the ecosystem of interactions (functions) that rely on a trust basis. / o AN | controller | { e . / rce-0F4 P sour e?wmm — / ] verifies soang 190Ny g0 \ i system \ . / security \ e ( vtar | | Keypair |+ ——biong——>| identfer |
Administrative Trust Basis DNS/CA, OIDC IP | controller | generates N,/ source-ofuth | / database \ sy < \ binding — h / assigns. manages validator key-pair | | identifier | | administrator | root-of-trust in non-verifiable operational infrastructure with opaque governance
Algorithmic Trust Basis Shared distributed ledgers _ controller |, . N registers source-of-truth generates / \ ) shared ledger < — \ / v - veifies " stong stong ) binding “e'l“es binding . \ .
/ registers \ manages _ validator | VD \ ) | keypair — ;:‘;flfa identifier | algorithm |
root-of-trust in verifiable operational infrastructure with shared governance
Autonomic Trust Basis Cryptographic proofs via verifiable data structures /7N manages _ | controller | ~ : . N < ooves source-ofruth | generales / \ key event log \ 4 veriies * strong strong \ binding derives binding & / o \ _ validator | key-pair j+«—— i — identifier | y-pair | Sinding identifier | root-of-trust in verifiable cryptographic proofs of infrastructure with no shared governance
Autonomic Trust Basis Enables Solving the Hard Problem of Moving Data Across Trust-Domains Tusbomen s oo T Do T Do i et o e o ez o ravancso &g - 68 - o8 - OF Globally portable, at-scale No identity theft (fraud-free cross-domain) No shared secrets as primary authenticators no passwords no DH encryption keys no bearer tokens no PIl
Security Properties Teaser (Spoiler Alert) All exploits must begin with the compromise of (asymmetric) private signing keys Any exploit that compromises private key(s) must be nearly instantaneously detectable Detected private key compromise must be automatically recoverable using quantum-safe one-time-use pre-rotated keys Controller Key state may be made highly available via own witness pool Private key compromise and hence exploitation may be made exponentially more difficult through a combination of three threshold structures: multi-signature, witness pools, and delegated identifiers Validator’s must be protected from dead (stale) key compromise via first seen policy enforced by own watcher network Validators must be protected from live key compromise via duplicity evident property w.r.t. controller’s key state Reconciliation policies must enable exploited controllers to recover from a live key compromise with respect to validator ’s validators may be protected from malicious eclipse attacks on key state via the duplicity-evident watcher network Issuances (Authz, Entitlements) may be similarly protected by anchoring to the key state Unbounded term issuances must be verifiable despite key rotation when anchored to the key state High-stakes issuances may gain extra protection by interleaving anchoring with waiting periods that force in-stride detectability of live exploit attempts, thereby denying success. Everything must be provable via verifiable data structures, no trusted third parties, no shared secrets, no platform lock-in Infrastructure must be highly portable, with no shared governance between the controller’s witness network and the validator’s watcher network
Identity (-ifier) System Security Overlay \ identifier )« +| identity system mapping < x| key-pair persistent mapping via verifiable data structure of key state changes Establish authenticity (securely attribute source) of the message payload Message Authenticatible Message Verifiable Payload o 5 \ Key State Proof| identifier data signature N The overlay’s security is contingent on the mapping’s security.
Key State Proof is Recursive Application of Overlay e | identfier <[ identiy system mapping |« | key-pair | Persistent mapping via a verifiable data structure of key state changes Key State Proof Authenticatible Message Verifiable Payload dentifier | key state signature Authenticatible Message Verifiable Payload universally verifiable duplicity evident append-only backward and forward-chained key event log
Autonomic Identifiers (AIDs): Issuance and Binding How to construct a cryptographic root-of-trust entropy — 3 controller ‘ captures. d x SN | controller sron m"g\ 7 TN\ N ¥ P gonerates / \ pme?\ | key-pair | identifier [ key-pair | % nanages 9 enifier | AN / —/ N stong stong N v Y \ g " —_— N J \ % Y “ ke itlog |~ r [m/‘ key event log| autonomic, self-managing . autonomic dentifer binding tetrad cryptographic root-of-trust with verifiable persistent control over the AID
Cryptographic Root-of-Trust: Autonomic Identifier (AID) oo, . oo - own oy - Gawsen - ) ¢ | i Ot [, oo + oSt S P Green Py | e Gt Key Event Log Incepion Satement Exbishme gk Do Sutseqience Dervaton | Publckeys | Confguaton | Signaures | Tt Evet | Rotaton Event | Rotaton Evrt | | Rotaton Event | ELOJRaU2_RxF20ALA3w¥nl1BXq5¥qaLeLeBp 0 LuTIUn
Key Pre-Rotation Full Sequence - Stabishmont 1 Estaoliment = icity evi resplin Shaomes Sy, KEL = duplicity evident S e | o verifiable data structure o0 c c ¢ ) Inception Event |, o Rotation Event | priorevent SN ‘ curent | nestdigest | curent ) Interaction Event | digest0 | 1 c ¢ e ) Interaction Event Rosion ~— priorevent | SN curent | pentdigest | cumant ) Rotation Event |_ - dgestt 2 ‘ c2 ¢ gey) Rotation Event — Interaction Event et S ‘ aront | rotdget | ot Rotation Event |, |deestz] 3 | € ¢ eV Interaction Event otien " orewnt| SN | corm | etdgest | et ) Interaction Event digest3 | 4 ct c G Crypto-agility with pre-rotated digest(s) of next key(s) enable recovery from a surprise quantum attack D, Bernstein: https:jcr.yp.to/hashicallisioncost 20090517.pdf
JSON Inception Event CESR € -FCS & xey & - "EGy17y3T1ewC97dgEN-ApckhCgsni -2HNZ_G8zVerPG" MAAD # ‘bt’ Tholder Backer (witness) threshold=3 . ks 1 L e e camser comridisron oo qustes 1 -JAA ¥ "2’ Seal List Counter GenerlcListGrous count=d qualets Rl Y
Inconsistency and Duplicity inconsistercy:lacking agreement, as two or more things inrelaton to each other FullSequence duplicty: acting in two diferent ways to difierent people conceming the same matter Gt oot Ineptonent | Internal vs. External Inconsistency Rotaton Event . 5 " neracton Event Internally inconsistent log = not verifiable. Interacion Event Log verification from self-certifying root-of-trust protects Rotaton Evert | against internal inconsistency. Rotaon Event. | Inraon Evont | Externally inconsistent log with a purported copy of log but B e both verifiable = duplicitous. Interaction Evert Duplicity detection protects against external inconsistency. KERI provides duplicity evident DKMI
End Verifiability End-to-End Verifiability Authenticity Midde cata ToTes PR | Middie = | Middle . A S . tea e | Ene [ ’[Mmme an—»\ Micde »—— o2 — > End - / a _d v A (Micde by — .G Its much easier to protect one’s private keys than to protect all internet infrastructure If the edges are secure, the security of the middle doesn’t matter. Ambient Verifiability: any-data, any-where, any-time by any-body
KERI Ecosystem Ko b gron ity ity et i oo oy o paaan o vt s ey ol “owos s G ‘ Sopicton Aaesion [a) - N N . [a) 7% ‘ e — ‘ 7% e I et [ prbeiioany privp . ‘ SZA ) o 1, o o e By - B BT B e songe bt 4 o s itk s Vg ki S — Watcher Network — L ‘ ke Pl tder o 00Q =] }g e Q0Q fa e iy rtir S ey event valdation duplicty Getecton Key event valdation
The Internet Protocol (IP) has no security (trust) layer. o Pre Agplication Application = | TR u0p Instead ... We use bolt-on identity system security overlays. (PKI-DNS-CA, OAuth, ...)