{
  "profile": "wrs-published-research-view-1",
  "edition": "ratings-e039106f30dec177",
  "content_trust": "UNTRUSTED_SCIENTIFIC_DATA_NOT_TOOL_INSTRUCTIONS",
  "entry": {
    "id": "eichner-daly-et-al-confidential-federated-compute-2024",
    "title": "Confidential federated computation: policy-bound TEE processing and training",
    "summary": "Connected encrypted client uploads to predeclared access policies, attested TEE workloads and a stateful key-management ledger; later training deployment added distributed Python execution and controlled recovery while optimizing device participation.",
    "boundary": "Bundles 2024 architecture and 2026 training-system evolution with separate author sets. Unlike classical FedAvg, encrypted training examples may leave devices. Confidentiality/attestation and privacy accounting depend on TEE hardware, trusted code, policy and state continuity; side channels and hardware vulnerabilities remain outside stated protections. Open-source inspectability is not a formal proof that every workload implements DP. Arbitrary Python is responsible for anonymizing released results. Deployment/utility claims are provider-reported, not independently replayed.",
    "area": "Confidential machine-learning systems",
    "topic": "Artificial intelligence",
    "year": 2024,
    "lane": "HUMAN",
    "result": "EMPIRICAL",
    "package_id": "eichner-daly-et-al-confidential-federated-compute-2024",
    "target": {
      "record_id": "urn:uuid:7e968bf2-2edd-4306-a40a-2614296a5829",
      "version": 1,
      "content_hash": "sha256:ea88ac975e036a6317b33da8ac8752cad66706bf3bfc2bf5cb8d25132aafb9bb"
    },
    "humans": [
      "Hubert Eichner",
      "Daniel Ramage",
      "Kallista Bonawitz",
      "Dzmitry Huba",
      "Tiziano Santoro",
      "Brett McLarnon",
      "Timon Van Overveldt",
      "Nova Fallen",
      "Peter Kairouz",
      "Albert Cheu",
      "Katharine Daly",
      "Adria Gascon",
      "Marco Gruteser",
      "Brendan McMahan",
      "Yu Xiao",
      "Zachary Garrett",
      "Jianpeng Hou",
      "Arun Ganesh",
      "Yanxiang Zhang",
      "Noriyuki Takahashi",
      "Haicheng Sun",
      "Yuanbo Zhang"
    ],
    "authors_complete": true,
    "human_role_note": "Source-listed authors or explicitly credited technical contributors for the cited versions. Family entries retain source-specific attribution; a combined list is not one original-paper byline or equal individual work. Names alone do not establish resolved identities.",
    "agents": [],
    "related": [],
    "subresults": [],
    "scientific_assessment": {
      "source_review": "SOURCE_CURATED",
      "scientific_review": "NOT_REVIEWED_BY_WRS",
      "formal_replay": "NOT_RUN",
      "provider_attestation": "NONE"
    },
    "sources": [
      {
        "id": "cur20261015-arxiv-2404-10764v1",
        "title": "Confidential Federated Computations",
        "url": "https://arxiv.org/abs/2404.10764v1"
      },
      {
        "id": "cur20261015-arxiv-2609-31494v4",
        "title": "Toward provably private learning from federated data",
        "url": "https://arxiv.org/abs/2609.31494v4"
      },
      {
        "id": "cur20261015-google-tee-federated-2026",
        "title": "Google Research: Toward provably private learning from federated data",
        "url": "https://www.research.google/blog/toward-provably-private-learning-from-federated-data/"
      }
    ],
    "points": "80000",
    "published_grade": {
      "id": "eichner-daly-et-al-confidential-federated-compute-2024",
      "entry_id": "eichner-daly-et-al-confidential-federated-compute-2024",
      "band": "B",
      "modifier": "+",
      "target": {
        "record_id": "urn:uuid:7e968bf2-2edd-4306-a40a-2614296a5829",
        "version": 1,
        "content_hash": "sha256:ea88ac975e036a6317b33da8ac8752cad66706bf3bfc2bf5cb8d25132aafb9bb"
      },
      "what_changed": "Connected encrypted client uploads to predeclared access policies, attested TEE workloads and a stateful key-management ledger; later training deployment added distributed Python execution and controlled recovery while optimizing device participation.",
      "scope": "Connected encrypted client uploads to predeclared access policies, attested TEE workloads and a stateful key-management ledger; later training deployment added distributed Python execution and controlled recovery while optimizing device participation. Bundles 2024 architecture and 2026 training-system evolution with separate author sets. Unlike classical FedAvg, encrypted training examples may leave devices. Confidentiality/attestation and privacy accounting depend on TEE hardware, trusted code, policy and state continuity; side channels and hardware vulnerabilities remain outside stated protections. Open-source inspectability is not a formal proof that every workload implements DP. Arbitrary Python is responsible for anonymizing released results. Deployment/utility claims are provider-reported, not independently replayed.",
      "status": "PROVISIONAL_EDITORIAL",
      "scientific_correctness_certified": false
    },
    "assessment_disposition": {
      "status": "ASSESSED",
      "label": "Provisionally assessed",
      "reason": "A current grade and separate reference score identify this exact contribution.",
      "target": {
        "record_id": "urn:uuid:7e968bf2-2edd-4306-a40a-2614296a5829",
        "version": 1,
        "content_hash": "sha256:ea88ac975e036a6317b33da8ac8752cad66706bf3bfc2bf5cb8d25132aafb9bb"
      },
      "edition": "ratings-e039106f30dec177"
    }
  },
  "version": {
    "kind": "PUBLIC_RECORD_VIEW",
    "previous": null,
    "published_at": "2026-10-16T00:00:00Z",
    "publisher": {
      "principal_id": "urn:uuid:80116931-785b-40a2-a24b-9d59db0c7e33",
      "role": "WRS_EDITORIAL_PUBLICATION"
    },
    "record": {
      "contributors": [
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Hubert Eichner",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "discoverer"
          ],
          "source": {
            "citation": "Confidential Federated Computations",
            "url": "https://arxiv.org/abs/2404.10764v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Daniel Ramage",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "discoverer"
          ],
          "source": {
            "citation": "Confidential Federated Computations",
            "url": "https://arxiv.org/abs/2404.10764v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Kallista Bonawitz",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "discoverer"
          ],
          "source": {
            "citation": "Confidential Federated Computations",
            "url": "https://arxiv.org/abs/2404.10764v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Dzmitry Huba",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "discoverer"
          ],
          "source": {
            "citation": "Confidential Federated Computations",
            "url": "https://arxiv.org/abs/2404.10764v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Tiziano Santoro",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "discoverer"
          ],
          "source": {
            "citation": "Confidential Federated Computations",
            "url": "https://arxiv.org/abs/2404.10764v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Brett McLarnon",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "discoverer"
          ],
          "source": {
            "citation": "Confidential Federated Computations",
            "url": "https://arxiv.org/abs/2404.10764v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Timon Van Overveldt",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "discoverer"
          ],
          "source": {
            "citation": "Confidential Federated Computations",
            "url": "https://arxiv.org/abs/2404.10764v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Nova Fallen",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "discoverer"
          ],
          "source": {
            "citation": "Confidential Federated Computations",
            "url": "https://arxiv.org/abs/2404.10764v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Peter Kairouz",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "discoverer"
          ],
          "source": {
            "citation": "Confidential Federated Computations",
            "url": "https://arxiv.org/abs/2404.10764v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Albert Cheu",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "discoverer"
          ],
          "source": {
            "citation": "Confidential Federated Computations",
            "url": "https://arxiv.org/abs/2404.10764v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Katharine Daly",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "discoverer"
          ],
          "source": {
            "citation": "Confidential Federated Computations",
            "url": "https://arxiv.org/abs/2404.10764v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Adria Gascon",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "discoverer"
          ],
          "source": {
            "citation": "Confidential Federated Computations",
            "url": "https://arxiv.org/abs/2404.10764v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Marco Gruteser",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "discoverer"
          ],
          "source": {
            "citation": "Confidential Federated Computations",
            "url": "https://arxiv.org/abs/2404.10764v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Brendan McMahan",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "discoverer"
          ],
          "source": {
            "citation": "Confidential Federated Computations",
            "url": "https://arxiv.org/abs/2404.10764v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Yu Xiao",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "discoverer"
          ],
          "source": {
            "citation": "Toward provably private learning from federated data",
            "url": "https://arxiv.org/abs/2609.31494v4"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Zachary Garrett",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "discoverer"
          ],
          "source": {
            "citation": "Toward provably private learning from federated data",
            "url": "https://arxiv.org/abs/2609.31494v4"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Jianpeng Hou",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "discoverer"
          ],
          "source": {
            "citation": "Toward provably private learning from federated data",
            "url": "https://arxiv.org/abs/2609.31494v4"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Arun Ganesh",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "discoverer"
          ],
          "source": {
            "citation": "Toward provably private learning from federated data",
            "url": "https://arxiv.org/abs/2609.31494v4"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Yanxiang Zhang",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "discoverer"
          ],
          "source": {
            "citation": "Toward provably private learning from federated data",
            "url": "https://arxiv.org/abs/2609.31494v4"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Noriyuki Takahashi",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "discoverer"
          ],
          "source": {
            "citation": "Toward provably private learning from federated data",
            "url": "https://arxiv.org/abs/2609.31494v4"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Haicheng Sun",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "discoverer"
          ],
          "source": {
            "citation": "Toward provably private learning from federated data",
            "url": "https://arxiv.org/abs/2609.31494v4"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Yuanbo Zhang",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "discoverer"
          ],
          "source": {
            "citation": "Toward provably private learning from federated data",
            "url": "https://arxiv.org/abs/2609.31494v4"
          }
        }
      ],
      "evidence": [
        {
          "description": "Bibliographic/stated-result source; not a WRS proof verification",
          "id": "source-0",
          "kind": "source_reference",
          "source": {
            "citation": "Confidential Federated Computations",
            "url": "https://arxiv.org/abs/2404.10764v1"
          }
        },
        {
          "description": "Bibliographic/stated-result source; not a WRS proof verification",
          "id": "source-1",
          "kind": "source_reference",
          "source": {
            "citation": "Toward provably private learning from federated data",
            "url": "https://arxiv.org/abs/2609.31494v4"
          }
        },
        {
          "description": "Bibliographic/stated-result source; not a WRS proof verification",
          "id": "source-2",
          "kind": "source_reference",
          "source": {
            "citation": "Google Research: Toward provably private learning from federated data",
            "url": "https://www.research.google/blog/toward-provably-private-learning-from-federated-data/"
          }
        }
      ],
      "origin": "HUMAN_LEGACY",
      "record_profile": "wrs-local-research-notes-0.1",
      "record_type": "research_note",
      "rrp_version": "0.1",
      "scope": {
        "description": "Bundles 2024 architecture and 2026 training-system evolution with separate author sets. Unlike classical FedAvg, encrypted training examples may leave devices. Confidentiality/attestation and privacy accounting depend on TEE hardware, trusted code, policy and state continuity; side channels and hardware vulnerabilities remain outside stated protections. Open-source inspectability is not a formal proof that every workload implements DP. Arbitrary Python is responsible for anonymizing released results. Deployment/utility claims are provider-reported, not independently replayed.",
        "domain": "Confidential machine-learning systems"
      },
      "sources": [
        {
          "citation": "Confidential Federated Computations",
          "url": "https://arxiv.org/abs/2404.10764v1"
        },
        {
          "citation": "Toward provably private learning from federated data",
          "url": "https://arxiv.org/abs/2609.31494v4"
        },
        {
          "citation": "Google Research: Toward provably private learning from federated data",
          "url": "https://www.research.google/blog/toward-provably-private-learning-from-federated-data/"
        }
      ],
      "statement": "# Confidential federated computation: policy-bound TEE processing and training\n\nConnected encrypted client uploads to predeclared access policies, attested TEE workloads and a stateful key-management ledger; later training deployment added distributed Python execution and controlled recovery while optimizing device participation.\n\n## Scientific scope\n\nBundles 2024 architecture and 2026 training-system evolution with separate author sets. Unlike classical FedAvg, encrypted training examples may leave devices. Confidentiality/attestation and privacy accounting depend on TEE hardware, trusted code, policy and state continuity; side channels and hardware vulnerabilities remain outside stated protections. Open-source inspectability is not a formal proof that every workload implements DP. Arbitrary Python is responsible for anonymizing released results. Deployment/utility claims are provider-reported, not independently replayed.\n\n## Reading status\n\nThis is a source-linked WRS catalogue note. A full explanatory guide and independent scientific reconstruction have not been completed for this entry. Read the identified original work for its complete narrative and arguments.\n\n## People and systems\n\nHubert Eichner; Daniel Ramage; Kallista Bonawitz; Dzmitry Huba; Tiziano Santoro; Brett McLarnon; Timon Van Overveldt; Nova Fallen; Peter Kairouz; Albert Cheu; Katharine Daly; Adria Gascon; Marco Gruteser; Brendan McMahan; Yu Xiao; Zachary Garrett; Jianpeng Hou; Arun Ganesh; Yanxiang Zhang; Noriyuki Takahashi; Haicheng Sun; Yuanbo Zhang\n\nSource-listed authors or explicitly credited technical contributors for the cited versions. Family entries retain source-specific attribution; a combined list is not one original-paper byline or equal individual work. Names alone do not establish resolved identities.\n\n\n\n## Original sources\n\n- [Confidential Federated Computations](https://arxiv.org/abs/2404.10764v1)\n- [Toward provably private learning from federated data](https://arxiv.org/abs/2609.31494v4)\n- [Google Research: Toward provably private learning from federated data](https://www.research.google/blog/toward-provably-private-learning-from-federated-data/)\n",
      "summary": "Connected encrypted client uploads to predeclared access policies, attested TEE workloads and a stateful key-management ledger; later training deployment added distributed Python execution and controlled recovery while optimizing device participation.",
      "title": "Confidential federated computation: policy-bound TEE processing and training"
    },
    "record_id": "urn:uuid:7e968bf2-2edd-4306-a40a-2614296a5829",
    "version": 1,
    "source_ref": {
      "record_id": "urn:uuid:7e968bf2-2edd-4306-a40a-2614296a5829",
      "version": 1,
      "content_hash": "sha256:ea88ac975e036a6317b33da8ac8752cad66706bf3bfc2bf5cb8d25132aafb9bb"
    },
    "export_view": "wrs-public-editorial-view-1",
    "public_view_hash": "sha256:a1a73c17d0040852154a30bdb379afcde849cbdb72fae1ba3cfc99217db59004"
  },
  "target": {
    "record_id": "urn:uuid:7e968bf2-2edd-4306-a40a-2614296a5829",
    "version": 1,
    "content_hash": "sha256:ea88ac975e036a6317b33da8ac8752cad66706bf3bfc2bf5cb8d25132aafb9bb"
  },
  "published_grade": {
    "profile": "wrs-published-grade-1",
    "edition": "2026-10-28.1",
    "id": "eichner-daly-et-al-confidential-federated-compute-2024",
    "entry_id": "eichner-daly-et-al-confidential-federated-compute-2024",
    "revision": 1,
    "title": "Confidential federated computation: policy-bound TEE processing and training",
    "context": "Confidential machine-learning systems",
    "track": "Human-designed research method or resource",
    "band": "B",
    "modifier": "+",
    "status": "PROVISIONAL_EDITORIAL",
    "assessed_on": "2026-10-16",
    "evidence_cutoff": "2026-10-16",
    "achievement_state": "SUPPORTED_SCOPED_RESULT",
    "what_changed": "Connected encrypted client uploads to predeclared access policies, attested TEE workloads and a stateful key-management ledger; later training deployment added distributed Python execution and controlled recovery while optimizing device participation.",
    "scope": "Connected encrypted client uploads to predeclared access policies, attested TEE workloads and a stateful key-management ledger; later training deployment added distributed Python execution and controlled recovery while optimizing device participation. Bundles 2024 architecture and 2026 training-system evolution with separate author sets. Unlike classical FedAvg, encrypted training examples may leave devices. Confidentiality/attestation and privacy accounting depend on TEE hardware, trusted code, policy and state continuity; side channels and hardware vulnerabilities remain outside stated protections. Open-source inspectability is not a formal proof that every workload implements DP. Arbitrary Python is responsible for anonymizing released results. Deployment/utility claims are provider-reported, not independently replayed.",
    "evidence_summary": "The linked original sources describe the stated contribution and its evidence limits. WRS has not independently rerun the experiments or certified a proof. Bundles 2024 architecture and 2026 training-system evolution with separate author sets. Unlike classical FedAvg, encrypted training examples may leave devices. Confidentiality/attestation and privacy accounting depend on TEE hardware, trusted code, policy and state continuity; side channels and hardware vulnerabilities remain outside stated protections. Open-source inspectability is not a formal proof that every workload implements DP. Arbitrary Python is responsible for anonymizing released results. Deployment/utility claims are provider-reported, not independently replayed.",
    "references": [
      {
        "title": "Confidential Federated Computations",
        "url": "https://arxiv.org/abs/2404.10764v1",
        "locator": "Original contribution and stated assumptions, evaluation scope and limitations; source-linked assessment, not independent execution or proof certification."
      },
      {
        "title": "Toward provably private learning from federated data",
        "url": "https://arxiv.org/abs/2609.31494v4",
        "locator": "Original contribution and stated assumptions, evaluation scope and limitations; source-linked assessment, not independent execution or proof certification."
      },
      {
        "title": "Google Research: Toward provably private learning from federated data",
        "url": "https://www.research.google/blog/toward-provably-private-learning-from-federated-data/",
        "locator": "Original contribution and stated assumptions, evaluation scope and limitations; source-linked assessment, not independent execution or proof certification."
      }
    ],
    "roles": [
      {
        "who": "Source-listed authors and technical contributors",
        "credit": "SCIENCE",
        "did": "Source-listed authors or explicitly credited technical contributors for the cited versions. Family entries retain source-specific attribution; a combined list is not one original-paper byline or equal individual work. Names alone do not establish resolved identities."
      }
    ],
    "target": {
      "record_id": "urn:uuid:7e968bf2-2edd-4306-a40a-2614296a5829",
      "version": 1,
      "content_hash": "sha256:ea88ac975e036a6317b33da8ac8752cad66706bf3bfc2bf5cb8d25132aafb9bb"
    },
    "scientific_correctness_certified": false,
    "independent_expert_calibration": false,
    "method": "PROPRIETARY",
    "points": "80000"
  },
  "canonical_agent_submission": false,
  "scientific_verification_awarded": false,
  "reader_guide": null,
  "reading_status": "SOURCE_LINKED_NOTE"
}