{"ontology":"https://layer.re/machine/schema.json","epistemic_model":"evidence-backed-temporal-layer-graph","canonical":true,"type":"civilization_snapshot","snapshot":{"id":"civ-20261008-persistent-agent-principals","as_of":"2026-10-08 19:29:00.000","summary":"New Google Cloud evidence sharpens the agent stack from persistent software infrastructure into a more explicit organizational-principal architecture. Earlier in 2026, Agent Identity became a first-class, cryptographically attested cloud principal and later gained generally available identity APIs and authentication brokering. On October 8, Google introduced a persistent Gemini work agent with cross-device context, headless and long-running execution, multi-agent orchestration and coworker agents with dedicated identities and organizational resources. The combined evidence materially advances Agents, Identity and Permissions while leaving cross-provider portability, independent reliability, adoption breadth and universal delegated-authority semantics unresolved.","cross_layer_changes":["Agent identity is now supported by production evidence as a first-class principal rather than only inherited user context.","Agent authentication and credential brokering have become explicit infrastructure for scoped tool access.","Persistent agents are being packaged as organizational workers with durable context, separate identities and governed resources.","Identity and permissions controls increasingly form a shared control plane around long-running agent action."],"dependencies":["agent identity + permissions + software -> separately governed persistent agents","identity APIs + auth brokering -> scoped external tool access","persistent cloud context + models + orchestration -> long-running organizational agent workflows"],"constraints":["cross-provider identity portability","delegation and revocation semantics","long-horizon reliability","independent security validation","rollout and adoption breadth"],"emerging_capabilities":["first-class agent principals","agent-specific authentication brokering","persistent coworker-agent accounts","headless cross-surface agent access","multi-agent orchestration with separate identities","model-decoupled persistent agent context"],"possible_bottlenecks":["portable authorization standards","reliable revocation across external systems","independent long-horizon task evaluation","audit interoperability"],"uncertainties":["Google's product architecture does not establish broad ecosystem adoption.","Vendor claims about hours-or-days execution are not independent reliability measurements.","Coworker-agent identity does not establish portable cross-provider agent identity.","Procedural memory and self-written skills do not establish recursive model self-improvement."],"mapping_mode":"contemporaneous","reconstructed_at":null,"created_at":"2026-10-08T20:03:28.742Z","updated_at":"2026-10-08T20:03:28.742Z"},"claims":[{"id":"claim-agent-control-plane-convergence-20261008","statement":"Across Google's 2026 agent stack, identity, scoped authorization, credential brokering, audit attribution, gateway policy and sandbox boundaries are converging around persistent agent action.","epistemic_status":"supported_inference","confidence":0.95,"civilization_snapshot_id":"civ-20261008-persistent-agent-principals","layer_state_snapshot_id":null,"competing_interpretation":"These controls may remain a vertically integrated Google-specific stack rather than a general industry architecture.","uncertainty":"Cross-provider convergence and bypass resistance require independent evidence.","mapping_mode":"contemporaneous","reconstructed_at":null,"created_at":"2026-10-08T20:03:28.744Z","updated_at":"2026-10-08T20:03:28.744Z"},{"id":"claim-google-agent-auth-broker-ga-20261008","statement":"By August 22, 2026, Google Cloud had made Agent Identity APIs and the Agent Identity auth manager generally available for agent identity management and OAuth or API-key credential brokering.","epistemic_status":"supported_inference","confidence":0.99,"civilization_snapshot_id":"civ-20261008-persistent-agent-principals","layer_state_snapshot_id":"state-permissions-20261008-agent-control-plane","competing_interpretation":"General availability does not establish broad customer adoption or cross-provider interoperability.","uncertainty":"Operational adoption and independent security properties remain unverified.","mapping_mode":"contemporaneous","reconstructed_at":null,"created_at":"2026-10-08T20:03:28.743Z","updated_at":"2026-10-08T20:03:28.743Z"},{"id":"claim-google-first-class-agent-identity-20261008","statement":"Google Cloud has production evidence for agent identity as a first-class principal distinct from human identities and generic service accounts, with cryptographic attestation and agent-specific IAM treatment.","epistemic_status":"supported_inference","confidence":0.99,"civilization_snapshot_id":"civ-20261008-persistent-agent-principals","layer_state_snapshot_id":"state-identity-20261008-agent-principals","competing_interpretation":"The capability may remain largely provider-specific rather than representing an ecosystem-wide identity standard.","uncertainty":"Cross-provider adoption, federation and independent validation remain limited.","mapping_mode":"contemporaneous","reconstructed_at":null,"created_at":"2026-10-08T20:03:28.743Z","updated_at":"2026-10-08T20:03:28.743Z"},{"id":"claim-google-persistent-coworker-agent-20261008","statement":"Google's October 8 Gemini agent architecture combines persistent cloud context and execution with multi-agent orchestration and coworker agents that have dedicated identities and organizational resources with scoped access.","epistemic_status":"supported_inference","confidence":0.97,"civilization_snapshot_id":"civ-20261008-persistent-agent-principals","layer_state_snapshot_id":"state-agents-20261008-persistent-principals","competing_interpretation":"The announcement may represent product architecture ahead of proven broad operational reliability or adoption.","uncertainty":"Independent multi-day completion rates, rollout breadth and human oversight frequency are not established.","mapping_mode":"contemporaneous","reconstructed_at":null,"created_at":"2026-10-08T20:03:28.743Z","updated_at":"2026-10-08T20:03:28.743Z"}],"layer_states":[{"id":"state-agents-20261008-persistent-principals","layer_id":"layer-agents","as_of":"2026-10-08 19:29:00.000","summary":"Persistent agents have moved beyond supporting state and application infrastructure toward an explicit organizational-worker architecture. Google documents cloud-persistent memory and context across surfaces, headless access, long-running delegated work, multi-agent orchestration, model choice separated from the persistent agent, and coworker agents with persistent roles and dedicated identities and resources. This is product-architecture evidence, not proof of reliable unsupervised multi-day performance or broad adoption.","dependencies":["layer-models","layer-compute","layer-software","layer-identity","layer-permissions","layer-human"],"constraints":["long-horizon reliability","rollout and adoption breadth","human oversight","cross-provider portability","cost and operational governance"],"emerging_capabilities":["persistent cross-device agent context","headless agent access","long-running delegated tasks","dynamic sub-agent orchestration","model-decoupled agent substrate","persistent coworker-agent roles"],"underdeveloped_components":["independent long-horizon task telemetry","portable agent state standards","cross-provider identity and delegation semantics"],"uncertainties":["independent reliability of hours-or-days execution","availability scope of all coworker-agent capabilities","degree of human intervention in real deployments"],"confidence":0.95,"previous_snapshot_id":"state-agents-20261004-state-infra","mapping_mode":"contemporaneous","reconstructed_at":null,"created_at":"2026-10-08T20:03:28.741Z","updated_at":"2026-10-08T20:03:28.741Z","layer_slug":"agents","layer_name":"Agents"},{"id":"state-identity-20261008-agent-principals","layer_id":"layer-identity","as_of":"2026-10-08 19:29:00.000","summary":"Agent identity now has strong production evidence as a first-class principal rather than only a borrowed user context. Google Cloud provides cryptographically attested, lifecycle-bound Agent Identity distinct from human identities and generic service accounts, with GA APIs and authentication brokering; the Gemini coworker architecture then gives persistent organizational agents their own accounts, email, storage and directory presence. Cross-provider portability and universal identity semantics remain unresolved.","dependencies":["layer-agents","layer-permissions","layer-software","layer-human"],"constraints":["cross-provider portability","principal-to-human delegation semantics","revocation and lifecycle consistency","external-system identity mapping"],"emerging_capabilities":["first-class agent principals","cryptographically attested agent identity","agent-specific IAM policy","persistent organizational agent accounts","identity-bound audit attribution"],"underdeveloped_components":["portable agent identity standards","cross-cloud principal federation","independent validation of lifecycle and revocation behavior"],"uncertainties":["how widely first-class agent principals are adopted outside Google Cloud","how agent identity maps across third-party systems and providers"],"confidence":0.98,"previous_snapshot_id":"state-identity-20261004-agent-app-context","mapping_mode":"contemporaneous","reconstructed_at":null,"created_at":"2026-10-08T20:03:28.741Z","updated_at":"2026-10-08T20:03:28.741Z","layer_slug":"identity","layer_name":"Identity"},{"id":"state-permissions-20261008-agent-control-plane","layer_id":"layer-permissions","as_of":"2026-10-08 19:29:00.000","summary":"Permission systems now have stronger hyperscaler evidence of an integrated agent control plane. Google combines agent-specific principals, allow/deny IAM policy, GA authentication brokering, scoped organizational access, per-agent audit attribution, gateway enforcement and sandbox boundaries. This strengthens the existing runtime-gateway pattern without establishing universal, portable delegated-permission semantics or independently proven bypass resistance.","dependencies":["layer-identity","layer-agents","layer-software","layer-human"],"constraints":["provider-specific semantics","cross-system delegation","revocation portability","policy-bypass resistance","independent security validation"],"emerging_capabilities":["agent-specific least-privilege policy","credential brokering for external tools","per-agent audit attribution","scoped coworker-agent access","gateway and sandbox enforcement"],"underdeveloped_components":["cross-provider delegated authorization standards","portable revocation and audit semantics","independent bypass testing"],"uncertainties":["how consistently policies propagate across third-party systems","real-world enforcement robustness under adversarial conditions"],"confidence":0.96,"previous_snapshot_id":"state-permissions-20261007-runtime-gateways","mapping_mode":"contemporaneous","reconstructed_at":null,"created_at":"2026-10-08T20:03:28.741Z","updated_at":"2026-10-08T20:03:28.741Z","layer_slug":"permissions","layer_name":"Permissions"}]}