{"resourceId":"register-anthropic-mythos-cisco-ibm-products-existing-licenses-20261109","versions":[{"version":"external-a68e3865eed260fe12f73def041482289bf38307385060e074ce5d10242185d9","resource":{"id":"register-anthropic-mythos-cisco-ibm-products-existing-licenses-20261109","title":"The Register: Cisco exposes Mythos functionality inside AI Defense under existing licenses without giving customers the model; IBM follows with Concert","organization":"The Register (reporting on Cisco, Anthropic and IBM)","sector":"AI security tooling and frontier-model access controls; cross-SLED applicability","geography":"Global (Cisco Live Melbourne announcement); no U.S. public-sector detail","publishedAt":"November 10, 2026 00:00 UTC (November 9, 2026 evening U.S. Central)","publicationDate":"2026-11-09","eventDate":"2026-11-09","sourceName":"The Register","sourceLabel":"Independent technology reporting based on vendor statements","sourceUrl":"https://www.theregister.com/ai-and-ml/2026/11/10/anthropics-super-bug-finding-mythos-model-is-going-into-cisco-and-ibm-products/5308616","evidenceClass":"independent-reporting","outcomeClass":"emerging","topics":["data-security","governance-procurement","developers-agents"],"finding":"The Register reports that Cisco AI Defense red teaming will be able to call Anthropic's Mythos to test customer setups, and that Cisco will offer it under users' existing licenses at no additional cost. It frames Mythos as a model whose distribution has been controlled (Project Glasswing) because it could help attackers, notes the Cisco deal exposes functionality without model access, and reports IBM announced a similar Concert integration hours later. Anthropic says demand has created a long queue of buyers.","sledRelevance":"Interpretation: if Mythos-backed testing arrives under existing AI Defense licenses, SLED customers with AI Defense may gain access without a new procurement, which affects budget and contract review. It also signals that frontier cyber-capable models will reach public-sector environments through vendor products before agencies set their own policies.","evidence":"Reporting with on-record quotes from Cisco SVP Jeff Schultz and Anthropic APAC head of applied AI Natalie Mead; no hands-on testing, metrics or customer evidence. Anthropic declined to say whether similar third-party deals exist.","architectureImplications":"The model stays behind Cisco's AI Defense service; customers interact with structured test jobs, so integration is via AI Defense, not direct model APIs.","governanceImplications":"No-additional-cost inclusion may bypass procurement review that would normally assess a new AI subprocessor; agencies should check whether Anthropic becomes a processor of test data under existing Cisco terms.","securityPrivacyImplications":"Dual-use capability: controls rest on Cisco's wrapper and Anthropic program safeguards, which agencies cannot independently inspect.","caveats":"Pricing statement ('without additional cost') is vendor-stated and may change; availability is still limited to select enterprises per Cisco. The article's Mythos 5.2 pricing details concern Anthropic's direct API, not Cisco's product.","streamIds":["cisco"],"roles":{"sales":"Sales framing: the licensing angle matters more than the model name. The Register reports Cisco intends Mythos-backed testing under existing AI Defense licenses at no extra cost, so current AI Defense customers in SLED may see new capability without a new purchase. Use it to open a conversation about AI-agent testing governance and subprocessor review, not to claim exclusive or proven capability; access remains limited and pricing is vendor-stated.","engineering":"Engineering framing: there is no model endpoint to integrate. Mythos sits behind AI Defense's structured red-teaming interface, so engineering work is target registration, network access for testing, and log handling. Ask Cisco to document which data from test interactions is processed by Anthropic, where it is processed, and how long it is retained, since the Register piece does not address any of this.","delivery":"Delivery framing: plan for a capability that may switch on in a tenant without a project. Delivery leads should add a change-control and subprocessor review step so new frontier-model features in existing licenses are approved before use, and maintain a test authorization register. Track availability announcements, since Cisco still lists the feature as coming soon for select enterprises."},"roleDetails":{"sales":{"customerProblem":"Agencies may receive frontier-model capabilities through existing licenses without a procurement or privacy review designed for them.","stakeholders":["CISO","Procurement/contracts officer","Privacy officer","AI governance committee"],"discoveryQuestions":["Does your AI Defense contract cover new AI subprocessors automatically?","Who approves enabling new AI features in existing SaaS tenants?","Would your policy allow a frontier cyber-capable model to test your systems via a vendor?"],"valueHypothesis":"Existing AI Defense customers could evaluate advanced red teaming with lower procurement friction, if governance review keeps pace.","potentialEngagement":"A short contract and subprocessor review of AI Defense terms paired with a test-authorization policy draft.","unsupportedClaims":["That the feature is free in all contracts or regions","That Mythos finds more vulnerabilities than other tools","That public-sector tenants are eligible"]},"engineering":{"fit":"Relevant only where AI Defense is already deployed and AI apps or agents are exposed to testing.","architectureIntegration":"Through AI Defense's structured red-team jobs; no direct Mythos API access.","prerequisites":["Existing AI Defense tenant","Registered test targets","Documented data flows for test traffic"],"constraints":["Limited availability","Vendor-controlled safeguards not independently inspectable","Jurisdictional data-processing terms unknown"],"security":"Require documentation of Anthropic's role as processor, logging and retention for test interactions, and the ability to disable the feature tenant-wide.","proofOfValue":["Confirm in writing which contract terms cover Mythos processing","Verify a tenant-level control exists to enable or disable Mythos-backed tests"]},"delivery":{"work":"Governance updates so licensed AI features introduced mid-contract are reviewed before enablement.","dependencies":["Cisco contract documentation","Agency privacy and procurement policies"],"ownership":"Agency CISO with procurement and privacy office.","skillsAdoption":"Contract and privacy staff need familiarity with AI subprocessor and dual-use model risks.","governanceCheckpoints":["Subprocessor review before enablement","AI governance committee sign-off for dual-use testing tools"],"acceptanceCriteria":["Proposed: documented approval exists before any Mythos-backed test runs","Proposed: subprocessor list updated within 30 days of feature availability"],"risks":["Feature enabled without review","Terms change after enablement","Assumed no-cost inclusion proves incorrect"]}},"retrievedAt":"2026-11-10T04:09:38Z","enrichedAt":"2026-11-10T04:15:16Z","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"Limited direct accessibility content in the source; workforce implication is the need for staff able to configure, interpret and act on AI security or cost tooling outputs (interpretation).","procurementImplications":"Interpretation: check whether new AI features arrive under existing Cisco/Splunk contracts, whether new subprocessors (e.g., Anthropic) are introduced, and whether public-sector authorizations cover them before enablement.","operatingModelImplications":"Interpretation: security, observability and finance teams need a shared operating loop (test, observe cost/quality, remediate) with named owners; the source does not describe a public-sector operating model.","sourceVerification":{"openedUrl":"https://www.theregister.com/ai-and-ml/2026/11/10/anthropics-super-bug-finding-mythos-model-is-going-into-cisco-and-ibm-products/5308616","referenceExcerpt":"The Cisco deal exposes the functionality of Mythos to end-users, without letting them use the model itself.","promptVersion":"sled-research-v4.0","model":null,"basis":"agent-reported inspection"}}}]}