{ "schema_version": "1.2", "url": "https://trustbutveri.fyi/explorer/?mechanisms=M-0012,M-0009,M-0013&hide=io,training", "data_generated": "2026-10-08", "definitions": { "methodology": "https://trustbutveri.fyi/about/methodology/", "readiness": "https://trustbutveri.fyi/about/readiness/", "filters": [ { "id": "prover", "label": "Prover", "question": "How far can the party being checked be trusted?", "options": [ { "value": "cooperative", "label": "Cooperative" }, { "value": "semi-trusted", "label": "Semi-trusted" }, { "value": "adversarial", "label": "Adversarial" } ], "rule": "Keeps mechanisms whose threat model holds against at least this prover. Adversarial is the strongest assumption.", "about": "The prover is the party being checked. Semi-trusted designs rely on part of its stack: usually the chip vendor's hardware root of trust, its firmware or counters, or its supply-chain records. Adversarial designs aim to hold even if it cheats wherever the checks allow, within their stated assumptions." }, { "id": "onsite", "label": "Verifier devices on site", "question": "May the verifier install its own hardware at the prover's sites?", "options": [ { "value": "no", "label": "Not allowed" } ], "rule": "\"Not allowed\" removes mechanisms that need a retrofit device, such as a network tap or a sealed sensor.", "about": "Some mechanisms need a device the verifier owns or trusts at the prover's facility, such as a network tap, a bandwidth limiter or a sealed sensor. Choose Not allowed when the setting rules that out. Inspectors are not covered." }, { "id": "coop", "label": "Prover cooperation", "question": "How much must the prover take part?", "options": [ { "value": "partial", "label": "Partial at most" }, { "value": "none", "label": "Not required" } ], "rule": "\"Partial at most\" removes mechanisms that need the prover's active participation. \"Not required\" keeps only those that work without it.", "about": "Required: the prover takes part, for example by logging requests, producing proofs or opening records. Partial: some access, such as installing a device. Not required: works from outside, such as satellite imagery." }, { "id": "chips", "label": "Chips", "question": "May the proposal depend on new chip designs?", "options": [ { "value": "existing", "label": "Existing chips only" } ], "rule": "\"Existing chips only\" removes mechanisms that need changes to future chip designs.", "about": "New chip features take years to reach a deployed fleet and cover only chips made after they ship. Mechanisms that use shipping features, such as trusted execution environments or performance counters, stay." }, { "id": "ready", "label": "Minimum readiness", "question": "How mature must each mechanism be?", "options": [ { "value": "R1", "label": "R1 Proposed" }, { "value": "R2", "label": "R2 Demonstrated" }, { "value": "R3", "label": "R3 In production" }, { "value": "R4", "label": "R4 Deployment-ready" } ], "rule": "Keeps mechanisms whose readiness level is at least this one.", "about": "A level describes the public evidence for a mechanism's stated use, not its cost or feasibility. R3 can still have open critical flaws." }, { "id": "tested", "label": "Attack testing", "question": "How hard has each mechanism been attacked in public?", "options": [ { "value": "analysis", "label": "Published analysis" }, { "value": "red-teamed", "label": "Red-teamed" }, { "value": "independent-red-team", "label": "Independent red-team" } ], "rule": "Keeps mechanisms whose strongest published attack testing is at least this.", "about": "The strongest published attempt to break the mechanism for its verification use: a security analysis, red-teaming by its developers or collaborators, or a red team independent of them." }, { "id": "hide", "label": "Keep hidden from the verifier", "question": "What must the verifier never see?", "options": [ { "value": "weights", "label": "Model weights" }, { "value": "io", "label": "Inputs and outputs" }, { "value": "training", "label": "Training data" } ], "rule": "Removes mechanisms that show the asset to the verifier. Conditional or unspecified exposure stays with a note and needs checking against the privacy requirement.", "about": "Model weights: the checked model's parameters. Inputs and outputs: the requests a deployed model serves and its responses. Training data: what a model was trained on. Each mechanism's exposure is the editors' reading of its record: shown, depends on the design (kept, with a note), hidden, not involved, or unspecified for a selected implementation. Code and configuration are not covered yet." } ], "exposure": "For model weights, inputs and outputs, and training data. This is the editors' reading of each mechanism's record (its threat model, how it works and its limitations), not a field of the record. Shown: the verifier sees it. Depends: on the design or variant, or the verifier sees only samples. Hidden: the verifier sees only commitments, hashes, proofs or results. Not involved: the record does not handle it. Unspecified: the selected implementation has no asset-specific assessment here.", "claim_status": { "addressed": "A mechanism in the proposal is aimed at this claim and is not excluded by the filters.", "partly-addressed": "Only supporting mechanisms, or mechanisms aimed at it that the filters exclude.", "unaddressed": "No mechanism in the proposal addresses this claim." }, "finding_scope": "Evidence scope describes where a finding was demonstrated; it does not establish applicability to every implementation in the mechanism family.", "claim_finding_scope": "open_critical_findings names active findings on the assessed records; open_critical_context names conditional family findings whose implementation applicability is unassessed.", "legacy_status": "The status field retains covered/partial/none for compatibility. It names claim links, never successful verification. Use claim_status and status_label for presentation." }, "filters": { "prover": "", "onsite": "", "coop": "", "chips": "", "ready": "", "tested": "", "hide": [ "io", "training" ] }, "mechanisms_passing_filters": 23, "claims": [], "mechanisms": [ { "id": "M-0012", "title": "Model identity attestation", "url": "https://trustbutveri.fyi/mechanisms/model-identity-attestation/", "assessment_record": { "id": "M-0012", "title": "Model identity attestation", "url": "https://trustbutveri.fyi/mechanisms/model-identity-attestation/" }, "selected_implementation": null, "readiness": { "level": "R3", "scope": "showing users that a service runs the declared model weights", "confidence": "medium", "evidence": [ "S-0013", "S-0012", "S-0015", "S-1206", "S-1207", "S-1208", "S-1209", "S-1202", "S-1507" ] }, "assessed_properties": { "threat_model": "semi-trusted", "hardware_requirement": "existing-features", "prover_cooperation": "required", "adversarial_evaluation": "independent-red-team" }, "claims": [], "exposure": { "weights": "partial", "io": "partial", "training": "none", "note": "The enclave route shows only hashes; the recomputation route gives the verifier the weights and the sampled requests and responses." }, "family_finding_context": [], "filter_issues": [ { "filter": "hide", "level": "note", "short": "may show inputs and outputs", "text": "May show inputs and outputs, depending on the design. The enclave route shows only hashes; the recomputation route gives the verifier the weights and the sampled requests and responses." } ] }, { "id": "M-0009", "title": "Hardware-enabled guarantees (flexHEG) and guarantee processors", "url": "https://trustbutveri.fyi/mechanisms/flexheg-guarantee-processors/", "assessment_record": { "id": "M-0009", "title": "Hardware-enabled guarantees (flexHEG) and guarantee processors", "url": "https://trustbutveri.fyi/mechanisms/flexheg-guarantee-processors/" }, "selected_implementation": null, "readiness": { "level": "R1", "scope": "checking and enforcing training-compute limits on chips, against adversaries up to states", "confidence": "medium", "evidence": [ "S-0035", "S-1204", "S-1205", "S-0057", "S-0056", "S-0006" ] }, "assessed_properties": { "threat_model": "adversarial", "hardware_requirement": "new-chip", "prover_cooperation": "required", "adversarial_evaluation": "analysis" }, "claims": [], "exposure": { "weights": "hidden", "io": "hidden", "training": "hidden", "note": "The guarantee processor sees the chip's traffic inside a sealed enclosure and reports only whether rules were kept." }, "family_finding_context": [], "filter_issues": [] }, { "id": "M-0013", "title": "Network taps and certifiers", "url": "https://trustbutveri.fyi/mechanisms/network-taps-and-certifiers/", "assessment_record": { "id": "M-0013", "title": "Network taps and certifiers", "url": "https://trustbutveri.fyi/mechanisms/network-taps-and-certifiers/" }, "selected_implementation": null, "readiness": { "level": "R1", "scope": "committing a complete record of cluster traffic, so declared inference can be checked", "confidence": "medium", "evidence": [ "S-1300", "S-0031", "S-0018", "S-0067", "S-1312", "S-1007", "S-1319", "S-1320" ] }, "assessed_properties": { "threat_model": "adversarial", "hardware_requirement": "retrofit-device", "prover_cooperation": "required", "adversarial_evaluation": "analysis" }, "claims": [], "exposure": { "weights": "partial", "io": "partial", "training": "partial", "note": "Only hashes leave the site; records picked for a challenge are opened for replay at a verification facility." }, "family_finding_context": [], "filter_issues": [ { "filter": "hide", "level": "note", "short": "may show inputs and outputs", "text": "May show inputs and outputs, depending on the design. Only hashes leave the site; records picked for a challenge are opened for replay at a verification facility." }, { "filter": "hide", "level": "note", "short": "may show training data", "text": "May show training data, depending on the design. Only hashes leave the site; records picked for a challenge are opened for replay at a verification facility." } ] } ], "strengths": { "covered": [], "production": [ "M-0012" ], "adversarial": [ "M-0009", "M-0013" ], "noNewHardware": [ "M-0012" ], "mitigated": [ { "mech": "M-0012", "n": 2, "title": "Launch-state attestation does not by itself cover weights loaded later", "kind": "theoretical-argument", "severity": "significant", "status": "mitigated", "evidence_scope": null, "scope_note": null, "related_finding": null, "description": "Attestation measures launch state, and weights are read from disk after boot. A signature checked at load time does not stop a malicious hypervisor from altering the disk afterwards. Tinfoil reports mitigating this with dm-verity checks on every read. Unmeasured runtime configuration remains a general risk.", "response": null, "sources": [ "S-0013", "S-0014" ] }, { "mech": "M-0013", "n": 6, "title": "Verifier dictionary attacks on hashes", "kind": "theoretical-argument", "severity": "minor", "status": "mitigated", "evidence_scope": null, "scope_note": null, "related_finding": null, "description": "Hashes of very short outputs could be brute-forced by the verifier. The paper recommends hashing at least 5 tokens together, or at least 10 if the attacker filters for likely tokens.", "response": null, "sources": [ "S-1300" ] } ], "notCounted": [] }, "properties": { "covered": [], "production": [ "M-0012" ], "adversarial": [ "M-0009", "M-0013" ], "noNewHardware": [ "M-0012" ], "mitigated": [ { "mech": "M-0012", "n": 2, "title": "Launch-state attestation does not by itself cover weights loaded later", "kind": "theoretical-argument", "severity": "significant", "status": "mitigated", "evidence_scope": null, "scope_note": null, "related_finding": null, "description": "Attestation measures launch state, and weights are read from disk after boot. A signature checked at load time does not stop a malicious hypervisor from altering the disk afterwards. Tinfoil reports mitigating this with dm-verity checks on every read. Unmeasured runtime configuration remains a general risk.", "response": null, "sources": [ "S-0013", "S-0014" ] }, { "mech": "M-0013", "n": 6, "title": "Verifier dictionary attacks on hashes", "kind": "theoretical-argument", "severity": "minor", "status": "mitigated", "evidence_scope": null, "scope_note": null, "related_finding": null, "description": "Hashes of very short outputs could be brute-forced by the verifier. The paper recommends hashing at least 5 tokens together, or at least 10 if the attacker filters for likely tokens.", "response": null, "sources": [ "S-1300" ] } ], "notCounted": [] }, "attack_testing": [ { "id": "M-0012", "record": "M-0012", "evaluation": "independent-red-team", "in_setting": true }, { "id": "M-0009", "record": "M-0009", "evaluation": "analysis", "in_setting": true }, { "id": "M-0013", "record": "M-0013", "evaluation": "analysis", "in_setting": true } ], "selected_implementations": {}, "weaknesses": { "gaps": [], "excluded": [], "unlinked": [], "critical": [ { "mech": "M-0012", "n": 1, "title": "Underlying attestation can be forged or relayed", "kind": "demonstrated-attack", "severity": "critical", "status": "open", "evidence_scope": "inherited", "scope_note": "Critical for the enclave route against an operator with physical access to affected hardware, or control of an unpatched SEV-SNP hypervisor. It does not apply to the recomputation route. PAL*M excludes physical attacks, and Tinfoil acknowledges this boundary.", "related_finding": { "record": "M-0008", "flaw": 1 }, "description": "The enclave route inherits the platform-specific TEE attestation failures. Intel TDX forgery and H100 relay were demonstrated with physical access and host control. Battering RAM defeated AMD SEV-SNP attestation on DDR4 servers; RMPocalypse did so from malicious host software on platforms without AMD's fixes. These demonstrate failures of the trust roots, not of each model-commitment protocol.", "response": "The TEE.fail authors report that physical interposer attacks are outside Intel's and AMD's threat models. AMD reports fixes for RMPocalypse.", "sources": [ "S-1202", "S-1210", "S-1212", "S-1213", "S-1206", "S-0012" ], "helps": [ { "by": "M-0009", "how": "A tamper-protected enclosure around the chip is the proposed answer when the party that holds the hardware may attack it physically." } ] } ], "significant": [ { "mech": "M-0012", "n": 3, "title": "For private models, a user can confirm consistency but not content", "kind": "open-question", "severity": "significant", "status": "open", "evidence_scope": null, "scope_note": null, "related_finding": null, "description": "When weights are not published, users can check that the same root hash is served each time, but not what the model is. Pairing the hash with an attested evaluation, as in Attestable Audits, is one proposed remedy.", "response": null, "sources": [ "S-0013", "S-0009" ] }, { "mech": "M-0012", "n": 4, "title": "Recomputation depends on trusted logging and randomness, and its tolerance leaves a covert channel", "kind": "demonstrated-attack", "severity": "significant", "status": "open", "evidence_scope": null, "scope_note": null, "related_finding": null, "description": "The recomputation variant assumes that every input, output and seed is logged correctly, and that the attacker can neither predict nor manipulate which messages are sampled for verification. Legitimate nondeterminism concentrates at a few token positions, and slow leaks within the tolerated slack remain possible. An independent study showed that an adversary who controls the prompts roughly doubles the bits leaked per token, reducing the exfiltration slowdown from 146–254 times under benign prompts to 60–118 times. The attack targets the exfiltration bound, not the check that outputs match the declared model.", "response": null, "sources": [ "S-0015", "S-1507" ], "helps": [ { "by": "M-0013", "how": "Taps are proposed to copy and hash traffic on the monitored links, reducing reliance on the prover's own log. This still depends on the monitored boundary and trusted capture." }, { "by": "M-0002", "how": "Bit-exact inference would remove the numerical tolerance that leaves this channel." } ] }, { "mech": "M-0009", "n": 1, "title": "State attackers can likely defeat current secure enclosures", "kind": "theoretical-argument", "severity": "significant", "status": "open", "evidence_scope": null, "scope_note": null, "related_finding": null, "description": "The flexHEG authors write that \"nation-state attackers can likely compromise the best current secure enclosures\", and that the marginal cost of circumvention per device is hard to estimate. RAND similarly judges that anti-tamper measures \"would not be insurmountable for a determined and well-resourced adversary\", although they raise costs and can reveal tampering.", "response": null, "sources": [ "S-1204", "S-0057" ] }, { "mech": "M-0009", "n": 2, "title": "Firmware-only retrofits rely on Secure Boot, which fault injection can bypass", "kind": "theoretical-argument", "severity": "significant", "status": "open", "evidence_scope": null, "scope_note": null, "related_finding": null, "description": "Part II notes that the most common attack on Secure Boot replaces the firmware and applies a voltage glitch while the signature is being checked. It also notes that sophisticated actors may use microprobing or laser voltage probing to read key registers.", "response": null, "sources": [ "S-1204" ] }, { "mech": "M-0009", "n": 3, "title": "Many important rules cannot be checked on-chip", "kind": "theoretical-argument", "severity": "significant", "status": "open", "evidence_scope": null, "scope_note": null, "related_finding": null, "description": "Malicious intent \"is not a technical property observable on-chip\", and misuse depends on what is done with a computation's results. A guarantee processor cannot easily tell whether a network is the whole system or one expert in a mixture-of-experts system. Part III judges that a fully local ruleset \"may not be entirely feasible\" for the same reason.", "response": null, "sources": [ "S-0035", "S-1205" ] }, { "mech": "M-0009", "n": 4, "title": "FLOP accounting can be laundered through external data", "kind": "theoretical-argument", "severity": "significant", "status": "open", "evidence_scope": null, "scope_note": null, "related_finding": null, "description": "Results of earlier or parallel workloads could be hidden in the \"external data\" fed to a device, which would falsify the total FLOP count unless the inputs are explained or time delays are imposed.", "response": null, "sources": [ "S-1204" ] }, { "mech": "M-0009", "n": 5, "title": "Supply-chain diversion and hidden backdoors", "kind": "open-question", "severity": "significant", "status": "open", "evidence_scope": null, "scope_note": null, "related_finding": null, "description": "Components could be diverted before a guarantee processor is added, and backdoors could be introduced during design or manufacturing. Open-source designs and physical scans of randomly selected chips are proposed as countermeasures. Part III proposes international oversight of production and extensive testing of a random sample of finished devices.", "response": null, "sources": [ "S-1204", "S-1205" ], "helps": [ { "by": "M-0019", "how": "Records each chip's identity and owner from the fab onwards, which bears on diversion before a guarantee processor is fitted. It does not address hidden backdoors." } ] }, { "mech": "M-0009", "n": 6, "title": "Coverage stops at flexHEG-equipped chips", "kind": "open-question", "severity": "significant", "status": "open", "evidence_scope": null, "scope_note": null, "related_finding": null, "description": "Motivated actors will always be able to use some compute that is not flexHEG-equipped. Recalling existing consumer GPUs would likely be impractical, and reaching perfect coverage, or conclusively proving that no secret government data centres exist, would be \"practically quite difficult\".", "response": null, "sources": [ "S-0035", "S-1205" ], "helps": [ { "by": "M-0019", "how": "Accounts for which chips exist and who holds them." }, { "by": "M-0020", "how": "Looks for undeclared facilities that hold other chips." } ] }, { "mech": "M-0013", "n": 1, "title": "Output nondeterminism leaves covert capacity", "kind": "theoretical-argument", "severity": "significant", "status": "open", "evidence_scope": null, "scope_note": null, "related_finding": null, "description": "Hashing cannot remove information hidden in the outputs themselves. The Secure Gateway Device paper estimates that about 0.1 bit per token remains even with seed-synchronized replay checks. For a 200k-GPU inference cluster at full load (2,000 tokens per GPU per second), that is about 40 Mbit/s of covert egress, enough to move a 1 TB model in under three days. The paper names this the core remaining challenge and points to deterministic replay or active scrubbing of hardware-induced entropy. An independent study found that an adversary who chooses the prompts roughly doubles the bits leaked per token under Gumbel-based inference verification; see Bounding unexplained information in outputs.", "response": null, "sources": [ "S-1300", "S-1507" ], "helps": [ { "by": "M-0002", "how": "Deterministic replay is one of the two remedies the flaw's source names." }, { "by": "M-0024", "how": "Bounds the hidden information outputs can carry by measuring what the declared computation fails to predict." } ] }, { "mech": "M-0013", "n": 2, "title": "Some links cannot be passively tapped", "kind": "open-question", "severity": "significant", "status": "open", "evidence_scope": null, "scope_note": null, "related_finding": null, "description": "Cankaya notes that copper-connected scale-up domains (for example NVL72 racks and TPU v7 cubes) are much harder to tap than fibre, and that optical budgets make passive taps impractical on 400GBASE-SR8 multimode links. Amodo found no taps advertised for 53 GBaud links as of May 2026.", "response": null, "sources": [ "S-0031", "S-1310" ] }, { "mech": "M-0013", "n": 3, "title": "Encrypted fabrics hide plaintext from both parties", "kind": "open-question", "severity": "significant", "status": "open", "evidence_scope": null, "scope_note": null, "related_finding": null, "description": "Cankaya notes that with TEE-protected sessions whose keys are ephemeral and managed inside the TEE, neither the operator nor the manufacturer can recover session keys after the session, so tapped traffic could not be opened for recomputation. For other encrypted fabrics, the operator can retain keys.", "response": null, "sources": [ "S-0031" ] }, { "mech": "M-0013", "n": 4, "title": "Residual side channels in simple passive setups", "kind": "theoretical-argument", "severity": "significant", "status": "open", "evidence_scope": null, "scope_note": null, "related_finding": null, "description": "Amodo's analysis of its own tapped prototype lists unvalidated header fields, timing of permitted traffic and variation in response formatting as residual channels, and concludes that the passive tap must be replaced by an active one.", "response": null, "sources": [ "S-1312" ] }, { "mech": "M-0013", "n": 5, "title": "Completeness rests on physical monitoring left out of scope", "kind": "open-question", "severity": "significant", "status": "open", "evidence_scope": null, "scope_note": null, "related_finding": null, "description": "The Secure Gateway Device paper assumes the facility is physically monitored, and states that the whole architecture depends on the device being the only communication channel. It names radio emanation, power-line signalling and thermal channels as covert channels beyond that scope.", "response": null, "sources": [ "S-1300" ], "helps": [ { "by": "M-0022", "how": "Addresses the radio, power-line and thermal channels that network-level designs leave out." } ] } ], "criticalMechanisms": [ "M-0012" ], "significantMechanisms": [ "M-0012", "M-0009", "M-0013" ], "familyContext": [], "minor": 0, "minorFindings": [], "minorBy": [], "notDemonstrated": [ "M-0009", "M-0013" ], "newChip": [ "M-0009" ] }, "findings": [ { "mech": "M-0012", "record": "M-0012", "n": 1, "title": "Underlying attestation can be forged or relayed", "kind": "demonstrated-attack", "severity": "critical", "status": "open", "evidence_scope": "inherited", "scope_note": "Critical for the enclave route against an operator with physical access to affected hardware, or control of an unpatched SEV-SNP hypervisor. It does not apply to the recomputation route. PAL*M excludes physical attacks, and Tinfoil acknowledges this boundary.", "related_finding": { "record": "M-0008", "flaw": 1 }, "description": "The enclave route inherits the platform-specific TEE attestation failures. Intel TDX forgery and H100 relay were demonstrated with physical access and host control. Battering RAM defeated AMD SEV-SNP attestation on DDR4 servers; RMPocalypse did so from malicious host software on platforms without AMD's fixes. These demonstrate failures of the trust roots, not of each model-commitment protocol.", "response": "The TEE.fail authors report that physical interposer attacks are outside Intel's and AMD's threat models. AMD reports fixes for RMPocalypse.", "sources": [ "S-1202", "S-1210", "S-1212", "S-1213", "S-1206", "S-0012" ], "helps": [ { "by": "M-0009", "how": "A tamper-protected enclosure around the chip is the proposed answer when the party that holds the hardware may attack it physically." } ] }, { "mech": "M-0012", "record": "M-0012", "n": 2, "title": "Launch-state attestation does not by itself cover weights loaded later", "kind": "theoretical-argument", "severity": "significant", "status": "mitigated", "evidence_scope": null, "scope_note": null, "related_finding": null, "description": "Attestation measures launch state, and weights are read from disk after boot. A signature checked at load time does not stop a malicious hypervisor from altering the disk afterwards. Tinfoil reports mitigating this with dm-verity checks on every read. Unmeasured runtime configuration remains a general risk.", "response": null, "sources": [ "S-0013", "S-0014" ] }, { "mech": "M-0012", "record": "M-0012", "n": 3, "title": "For private models, a user can confirm consistency but not content", "kind": "open-question", "severity": "significant", "status": "open", "evidence_scope": null, "scope_note": null, "related_finding": null, "description": "When weights are not published, users can check that the same root hash is served each time, but not what the model is. Pairing the hash with an attested evaluation, as in Attestable Audits, is one proposed remedy.", "response": null, "sources": [ "S-0013", "S-0009" ] }, { "mech": "M-0012", "record": "M-0012", "n": 4, "title": "Recomputation depends on trusted logging and randomness, and its tolerance leaves a covert channel", "kind": "demonstrated-attack", "severity": "significant", "status": "open", "evidence_scope": null, "scope_note": null, "related_finding": null, "description": "The recomputation variant assumes that every input, output and seed is logged correctly, and that the attacker can neither predict nor manipulate which messages are sampled for verification. Legitimate nondeterminism concentrates at a few token positions, and slow leaks within the tolerated slack remain possible. An independent study showed that an adversary who controls the prompts roughly doubles the bits leaked per token, reducing the exfiltration slowdown from 146–254 times under benign prompts to 60–118 times. The attack targets the exfiltration bound, not the check that outputs match the declared model.", "response": null, "sources": [ "S-0015", "S-1507" ], "helps": [ { "by": "M-0013", "how": "Taps are proposed to copy and hash traffic on the monitored links, reducing reliance on the prover's own log. This still depends on the monitored boundary and trusted capture." }, { "by": "M-0002", "how": "Bit-exact inference would remove the numerical tolerance that leaves this channel." } ] }, { "mech": "M-0009", "record": "M-0009", "n": 1, "title": "State attackers can likely defeat current secure enclosures", "kind": "theoretical-argument", "severity": "significant", "status": "open", "evidence_scope": null, "scope_note": null, "related_finding": null, "description": "The flexHEG authors write that \"nation-state attackers can likely compromise the best current secure enclosures\", and that the marginal cost of circumvention per device is hard to estimate. RAND similarly judges that anti-tamper measures \"would not be insurmountable for a determined and well-resourced adversary\", although they raise costs and can reveal tampering.", "response": null, "sources": [ "S-1204", "S-0057" ] }, { "mech": "M-0009", "record": "M-0009", "n": 2, "title": "Firmware-only retrofits rely on Secure Boot, which fault injection can bypass", "kind": "theoretical-argument", "severity": "significant", "status": "open", "evidence_scope": null, "scope_note": null, "related_finding": null, "description": "Part II notes that the most common attack on Secure Boot replaces the firmware and applies a voltage glitch while the signature is being checked. It also notes that sophisticated actors may use microprobing or laser voltage probing to read key registers.", "response": null, "sources": [ "S-1204" ] }, { "mech": "M-0009", "record": "M-0009", "n": 3, "title": "Many important rules cannot be checked on-chip", "kind": "theoretical-argument", "severity": "significant", "status": "open", "evidence_scope": null, "scope_note": null, "related_finding": null, "description": "Malicious intent \"is not a technical property observable on-chip\", and misuse depends on what is done with a computation's results. A guarantee processor cannot easily tell whether a network is the whole system or one expert in a mixture-of-experts system. Part III judges that a fully local ruleset \"may not be entirely feasible\" for the same reason.", "response": null, "sources": [ "S-0035", "S-1205" ] }, { "mech": "M-0009", "record": "M-0009", "n": 4, "title": "FLOP accounting can be laundered through external data", "kind": "theoretical-argument", "severity": "significant", "status": "open", "evidence_scope": null, "scope_note": null, "related_finding": null, "description": "Results of earlier or parallel workloads could be hidden in the \"external data\" fed to a device, which would falsify the total FLOP count unless the inputs are explained or time delays are imposed.", "response": null, "sources": [ "S-1204" ] }, { "mech": "M-0009", "record": "M-0009", "n": 5, "title": "Supply-chain diversion and hidden backdoors", "kind": "open-question", "severity": "significant", "status": "open", "evidence_scope": null, "scope_note": null, "related_finding": null, "description": "Components could be diverted before a guarantee processor is added, and backdoors could be introduced during design or manufacturing. Open-source designs and physical scans of randomly selected chips are proposed as countermeasures. Part III proposes international oversight of production and extensive testing of a random sample of finished devices.", "response": null, "sources": [ "S-1204", "S-1205" ], "helps": [ { "by": "M-0019", "how": "Records each chip's identity and owner from the fab onwards, which bears on diversion before a guarantee processor is fitted. It does not address hidden backdoors." } ] }, { "mech": "M-0009", "record": "M-0009", "n": 6, "title": "Coverage stops at flexHEG-equipped chips", "kind": "open-question", "severity": "significant", "status": "open", "evidence_scope": null, "scope_note": null, "related_finding": null, "description": "Motivated actors will always be able to use some compute that is not flexHEG-equipped. Recalling existing consumer GPUs would likely be impractical, and reaching perfect coverage, or conclusively proving that no secret government data centres exist, would be \"practically quite difficult\".", "response": null, "sources": [ "S-0035", "S-1205" ], "helps": [ { "by": "M-0019", "how": "Accounts for which chips exist and who holds them." }, { "by": "M-0020", "how": "Looks for undeclared facilities that hold other chips." } ] }, { "mech": "M-0013", "record": "M-0013", "n": 1, "title": "Output nondeterminism leaves covert capacity", "kind": "theoretical-argument", "severity": "significant", "status": "open", "evidence_scope": null, "scope_note": null, "related_finding": null, "description": "Hashing cannot remove information hidden in the outputs themselves. The Secure Gateway Device paper estimates that about 0.1 bit per token remains even with seed-synchronized replay checks. For a 200k-GPU inference cluster at full load (2,000 tokens per GPU per second), that is about 40 Mbit/s of covert egress, enough to move a 1 TB model in under three days. The paper names this the core remaining challenge and points to deterministic replay or active scrubbing of hardware-induced entropy. An independent study found that an adversary who chooses the prompts roughly doubles the bits leaked per token under Gumbel-based inference verification; see Bounding unexplained information in outputs.", "response": null, "sources": [ "S-1300", "S-1507" ], "helps": [ { "by": "M-0002", "how": "Deterministic replay is one of the two remedies the flaw's source names." }, { "by": "M-0024", "how": "Bounds the hidden information outputs can carry by measuring what the declared computation fails to predict." } ] }, { "mech": "M-0013", "record": "M-0013", "n": 2, "title": "Some links cannot be passively tapped", "kind": "open-question", "severity": "significant", "status": "open", "evidence_scope": null, "scope_note": null, "related_finding": null, "description": "Cankaya notes that copper-connected scale-up domains (for example NVL72 racks and TPU v7 cubes) are much harder to tap than fibre, and that optical budgets make passive taps impractical on 400GBASE-SR8 multimode links. Amodo found no taps advertised for 53 GBaud links as of May 2026.", "response": null, "sources": [ "S-0031", "S-1310" ] }, { "mech": "M-0013", "record": "M-0013", "n": 3, "title": "Encrypted fabrics hide plaintext from both parties", "kind": "open-question", "severity": "significant", "status": "open", "evidence_scope": null, "scope_note": null, "related_finding": null, "description": "Cankaya notes that with TEE-protected sessions whose keys are ephemeral and managed inside the TEE, neither the operator nor the manufacturer can recover session keys after the session, so tapped traffic could not be opened for recomputation. For other encrypted fabrics, the operator can retain keys.", "response": null, "sources": [ "S-0031" ] }, { "mech": "M-0013", "record": "M-0013", "n": 4, "title": "Residual side channels in simple passive setups", "kind": "theoretical-argument", "severity": "significant", "status": "open", "evidence_scope": null, "scope_note": null, "related_finding": null, "description": "Amodo's analysis of its own tapped prototype lists unvalidated header fields, timing of permitted traffic and variation in response formatting as residual channels, and concludes that the passive tap must be replaced by an active one.", "response": null, "sources": [ "S-1312" ] }, { "mech": "M-0013", "record": "M-0013", "n": 5, "title": "Completeness rests on physical monitoring left out of scope", "kind": "open-question", "severity": "significant", "status": "open", "evidence_scope": null, "scope_note": null, "related_finding": null, "description": "The Secure Gateway Device paper assumes the facility is physically monitored, and states that the whole architecture depends on the device being the only communication channel. It names radio emanation, power-line signalling and thermal channels as covert channels beyond that scope.", "response": null, "sources": [ "S-1300" ], "helps": [ { "by": "M-0022", "how": "Addresses the radio, power-line and thermal channels that network-level designs leave out." } ] }, { "mech": "M-0013", "record": "M-0013", "n": 6, "title": "Verifier dictionary attacks on hashes", "kind": "theoretical-argument", "severity": "minor", "status": "mitigated", "evidence_scope": null, "scope_note": null, "related_finding": null, "description": "Hashes of very short outputs could be brute-forced by the verifier. The paper recommends hashing at least 5 tokens together, or at least 10 if the attacker filters for likely tokens.", "response": null, "sources": [ "S-1300" ] } ], "possible_additions": [ { "id": "M-0002", "title": "Deterministic and bit-exact inference", "url": "https://trustbutveri.fyi/mechanisms/deterministic-inference/", "readiness": "R3", "fits_filters": true, "filter_issues": [], "reasons": [ { "kind": "flaw", "mech": "M-0012", "n": 4, "title": "Recomputation depends on trusted logging and randomness, and its tolerance leaves a covert channel", "severity": "significant", "how": "Bit-exact inference would remove the numerical tolerance that leaves this channel." }, { "kind": "flaw", "mech": "M-0013", "n": 1, "title": "Output nondeterminism leaves covert capacity", "severity": "significant", "how": "Deterministic replay is one of the two remedies the flaw's source names." }, { "kind": "blocker", "mech": "M-0012", "text": "Numerical nondeterminism limits how tightly recomputation can pin down the model and sampling." }, { "kind": "blocker", "mech": "M-0013", "text": "Nondeterministic inference leaves covert capacity in outputs that hashing cannot remove." } ] }, { "id": "M-0024", "title": "Bounding unexplained information in outputs", "url": "https://trustbutveri.fyi/mechanisms/bounding-unexplained-information/", "readiness": "R2", "fits_filters": true, "filter_issues": [], "reasons": [ { "kind": "flaw", "mech": "M-0013", "n": 1, "title": "Output nondeterminism leaves covert capacity", "severity": "significant", "how": "Bounds the hidden information outputs can carry by measuring what the declared computation fails to predict." } ] }, { "id": "M-0019", "title": "Chip registries and manufacturing records", "url": "https://trustbutveri.fyi/mechanisms/chip-registries-and-manufacturing-records/", "readiness": "R1", "fits_filters": true, "filter_issues": [], "reasons": [ { "kind": "flaw", "mech": "M-0009", "n": 5, "title": "Supply-chain diversion and hidden backdoors", "severity": "significant", "how": "Records each chip's identity and owner from the fab onwards, which bears on diversion before a guarantee processor is fitted. It does not address hidden backdoors." }, { "kind": "flaw", "mech": "M-0009", "n": 6, "title": "Coverage stops at flexHEG-equipped chips", "severity": "significant", "how": "Accounts for which chips exist and who holds them." }, { "kind": "blocker", "mech": "M-0009", "text": "Governing all relevant chips depends on knowing where they are, through chip registries and detection of undeclared facilities." } ] }, { "id": "M-0020", "title": "Remote detection of data centres", "url": "https://trustbutveri.fyi/mechanisms/remote-detection-of-data-centres/", "readiness": "R1", "fits_filters": true, "filter_issues": [], "reasons": [ { "kind": "flaw", "mech": "M-0009", "n": 6, "title": "Coverage stops at flexHEG-equipped chips", "severity": "significant", "how": "Looks for undeclared facilities that hold other chips." } ] }, { "id": "M-0022", "title": "Side-channel suppression for isolated facilities", "url": "https://trustbutveri.fyi/mechanisms/side-channel-suppression/", "readiness": "R1", "fits_filters": true, "filter_issues": [], "reasons": [ { "kind": "flaw", "mech": "M-0013", "n": 5, "title": "Completeness rests on physical monitoring left out of scope", "severity": "significant", "how": "Addresses the radio, power-line and thermal channels that network-level designs leave out." }, { "kind": "blocker", "mech": "M-0013", "text": "Radio, power-line and thermal channels are not addressed by network-level designs." } ] }, { "id": "M-0008", "title": "TEE remote attestation for AI workloads", "url": "https://trustbutveri.fyi/mechanisms/tee-remote-attestation/", "readiness": "R3", "fits_filters": true, "filter_issues": [], "reasons": [ { "kind": "blocker", "mech": "M-0012", "text": "Attestation that resists physical attackers, for the enclave variant." }, { "kind": "prerequisite", "mech": "M-0009" } ] }, { "id": "M-0017", "title": "Tamper evidence for verifier devices", "url": "https://trustbutveri.fyi/mechanisms/tamper-evidence-for-verifier-devices/", "readiness": "R2", "fits_filters": true, "filter_issues": [], "reasons": [ { "kind": "blocker", "mech": "M-0009", "text": "State-level attackers who hold the hardware can likely compromise the best current secure enclosures." }, { "kind": "blocker", "mech": "M-0013", "text": "Taps and gateway devices need tamper-evident housing and physical monitoring so that traffic cannot bypass them." } ] }, { "id": "M-0001", "title": "Sampled inference recomputation", "url": "https://trustbutveri.fyi/mechanisms/sampled-inference-recomputation/", "readiness": "R3", "fits_filters": true, "filter_issues": [], "reasons": [ { "kind": "prerequisite", "mech": "M-0013" } ] } ], "goal": null, "design": null, "dependencies": { "prerequisites": [ { "id": "M-0008", "neededBy": [ "M-0009" ] }, { "id": "M-0019", "neededBy": [ "M-0009" ] }, { "id": "M-0001", "neededBy": [ "M-0013" ] }, { "id": "M-0002", "neededBy": [ "M-0013" ] }, { "id": "M-0017", "neededBy": [ "M-0013" ] }, { "id": "M-0022", "neededBy": [ "M-0013" ] } ], "shared": [ { "id": "M-0008", "by": [ "M-0012", "M-0009" ], "inProposal": false }, { "id": "M-0002", "by": [ "M-0012", "M-0013" ], "inProposal": false }, { "id": "M-0017", "by": [ "M-0009", "M-0013" ], "inProposal": false } ], "blockers": [ { "mech": "M-0012", "n": 1, "text": "Attestation that resists physical attackers, for the enclave variant.", "theme": "hardware-trust", "blocked_by": "M-0008", "sources": [ "S-1202" ], "inProposal": false }, { "mech": "M-0012", "n": 2, "text": "Numerical nondeterminism limits how tightly recomputation can pin down the model and sampling.", "theme": "protocol-soundness", "blocked_by": "M-0002", "sources": [ "S-0015" ], "inProposal": false }, { "mech": "M-0012", "n": 3, "text": "The recomputation variant needs the verifier to hold the declared weights.", "theme": "access-governance", "blocked_by": null, "sources": [ "S-0015" ], "inProposal": null }, { "mech": "M-0009", "n": 1, "text": "Integrated flexHEG needs substantial help from the accelerator manufacturer, and the authors estimate 3.7–7.9 years, from when the manufacturer starts work, for such hardware to displace other accelerators in frontier development.", "theme": "access-governance", "blocked_by": null, "sources": [ "S-1204" ], "inProposal": null }, { "mech": "M-0009", "n": 2, "text": "State-level attackers who hold the hardware can likely compromise the best current secure enclosures.", "theme": "hardware-trust", "blocked_by": "M-0017", "sources": [ "S-1204", "S-0057" ], "inProposal": false }, { "mech": "M-0009", "n": 3, "text": "Rival states would need to trust the design and manufacture of guarantee processors and enclosures, for example through open design, redundant processors from each side or oversight of production.", "theme": "hardware-trust", "blocked_by": null, "sources": [ "S-1205", "S-0035" ], "inProposal": null }, { "mech": "M-0009", "n": 4, "text": "Restricting future rule updates would need a formal language for rules, which the authors judge most likely infeasible for early flexHEG versions.", "theme": "protocol-soundness", "blocked_by": null, "sources": [ "S-0035" ], "inProposal": null }, { "mech": "M-0009", "n": 5, "text": "Governing all relevant chips depends on knowing where they are, through chip registries and detection of undeclared facilities.", "theme": "coverage-hidden-compute", "blocked_by": "M-0019", "sources": [ "S-1205" ], "inProposal": false }, { "mech": "M-0013", "n": 1, "text": "No complete verification tap has been demonstrated at production frontend link rates, and on the tested CPU no hash algorithm reached line rate with minimum-size frames.", "theme": "performance-compatibility", "blocked_by": null, "sources": [ "S-1311", "S-1310" ], "inProposal": null }, { "mech": "M-0013", "n": 2, "text": "Nondeterministic inference leaves covert capacity in outputs that hashing cannot remove.", "theme": "evidence-binding", "blocked_by": "M-0002", "sources": [ "S-1300" ], "inProposal": false }, { "mech": "M-0013", "n": 3, "text": "Taps and gateway devices need tamper-evident housing and physical monitoring so that traffic cannot bypass them.", "theme": "hardware-trust", "blocked_by": "M-0017", "sources": [ "S-1300", "S-0018" ], "inProposal": false }, { "mech": "M-0013", "n": 4, "text": "Radio, power-line and thermal channels are not addressed by network-level designs.", "theme": "coverage-hidden-compute", "blocked_by": "M-0022", "sources": [ "S-1300" ], "inProposal": false }, { "mech": "M-0013", "n": 5, "text": "Red-teaming by specialists is called for but has not been reported.", "theme": "adversarial-validation", "blocked_by": null, "sources": [ "S-1300" ], "inProposal": null } ] }, "exposure": { "weights": { "shown": [], "partial": [ "M-0012", "M-0013" ], "hidden": [ "M-0009" ], "none": [], "unknown": [] }, "io": { "shown": [], "partial": [ "M-0012", "M-0013" ], "hidden": [ "M-0009" ], "none": [], "unknown": [] }, "training": { "shown": [], "partial": [ "M-0013" ], "hidden": [ "M-0009" ], "none": [ "M-0012" ], "unknown": [] } }, "implementations": [ { "mechanism": "M-0012", "selected": null, "implementations": [ { "id": "I-0007", "title": "Attestable Audits", "url": "https://trustbutveri.fyi/implementations/attestable-audits/" }, { "id": "I-0022", "title": "PAL*M", "url": "https://trustbutveri.fyi/implementations/palm/" }, { "id": "I-0006", "title": "Tinfoil model identity (Modelwrap)", "url": "https://trustbutveri.fyi/implementations/tinfoil-model-identity/" } ] }, { "mechanism": "M-0009", "selected": null, "implementations": [] }, { "mechanism": "M-0013", "selected": null, "implementations": [ { "id": "I-0011", "title": "AI 2040 inference-only verification stack", "url": "https://trustbutveri.fyi/implementations/ai-2040-inference-only-verification-plan/" }, { "id": "I-0012", "title": "Low-trust AI compute verification system overview", "url": "https://trustbutveri.fyi/implementations/low-trust-compute-verification-system-overview/" }, { "id": "I-0008", "title": "SASH confidential network logger", "url": "https://trustbutveri.fyi/implementations/sash-confidential-network-logger/" } ] } ], "sources": [ { "id": "S-0013", "title": "How Tinfoil Proves Exactly What Model Is Running", "authors": "Tinfoil Team", "year": 2026, "url": "https://tinfoil.sh/blog/2026-02-03-proving-model-identity", "path": "/sources/tinfoil-proving-model-identity/" }, { "id": "S-0012", "title": "PAL*M: Property Attestation for Large Generative Models", "authors": "P. Chantasantitam et al.", "year": 2026, "url": "https://arxiv.org/abs/2601.16199", "path": "/sources/chantasantitam-palm/" }, { "id": "S-0015", "title": "Verifying LLM Inference to Detect Model Weight Exfiltration", "authors": "R. Rinberg et al.", "year": 2025, "url": "https://arxiv.org/abs/2511.02620", "path": "/sources/rinberg-verifying-llm-inference-weight-exfiltration/" }, { "id": "S-1206", "title": "A primer on secure enclaves", "authors": "Tinfoil", "year": 2026, "url": "https://docs.tinfoil.sh/verification/secure-enclave-primer", "path": "/sources/tinfoil-docs-secure-enclave-primer/" }, { "id": "S-1207", "title": "Backend infrastructure", "authors": "Tinfoil", "year": 2026, "url": "https://docs.tinfoil.sh/verification/attestation-architecture", "path": "/sources/tinfoil-docs-attestation-architecture/" }, { "id": "S-1208", "title": "How verification works in Tinfoil", "authors": "Tinfoil", "year": 2026, "url": "https://docs.tinfoil.sh/verification/verification-in-tinfoil", "path": "/sources/tinfoil-docs-verification-in-tinfoil/" }, { "id": "S-1209", "title": "modelwrap: Reproducible dm-verity read-only image of Huggingface models", "authors": "Tinfoil", "year": 2026, "url": "https://github.com/tinfoilsh/modelwrap", "path": "/sources/tinfoil-modelwrap-code/" }, { "id": "S-1202", "title": "TEE.fail: Breaking Trusted Execution Environments via DDR5 Memory Bus Interposition", "authors": "J. Chuang et al.", "year": 2026, "url": "https://tee.fail/", "path": "/sources/chuang-tee-fail/" }, { "id": "S-1507", "title": "Adversarial Entropy Inflation Against Gumbel-Based Inference Verification", "authors": "N. Kezins", "year": 2026, "url": "https://arxiv.org/abs/2608.23375", "path": "/sources/kezins-adversarial-entropy-inflation/" }, { "id": "S-1210", "title": "Battering RAM: Low-Cost Interposer Attacks on Confidential Computing via Dynamic Memory Aliasing", "authors": "J. De Meulemeester et al.", "year": 2026, "url": "https://batteringram.eu/", "path": "/sources/de-meulemeester-battering-ram/" }, { "id": "S-1212", "title": "RMPocalypse: How a Catch-22 Breaks AMD SEV-SNP", "authors": "B. Schlüter & S. Shinde", "year": 2025, "url": "https://rmpocalypse.github.io/", "path": "/sources/schluter-rmpocalypse/" }, { "id": "S-1213", "title": "SEV-SNP RMP Initialization Vulnerability (AMD-SB-3020)", "authors": "AMD", "year": 2025, "url": "https://www.amd.com/en/resources/product-security/bulletin/amd-sb-3020.html", "path": "/sources/amd-sb-3020-rmp-initialization/" }, { "id": "S-0014", "title": "On TEEs for Privacy-Preserving Monitoring in AI Governance", "authors": "Gloria Z", "year": 2026, "url": "https://techgov.intelligence.org/blog/on-tees-for-privacy-preserving-monitoring-in-ai-governance", "path": "/sources/zhao-tees-privacy-preserving-monitoring/" }, { "id": "S-0009", "title": "Attestable Audits: Verifiable AI Safety Benchmarks Using Trusted Execution Environments", "authors": "C. Schnabl et al.", "year": 2025, "url": "https://arxiv.org/abs/2506.23706", "path": "/sources/schnabl-attestable-audits/" }, { "id": "S-0035", "title": "Flexible Hardware-Enabled Guarantees for AI Compute", "authors": "J. Petrie et al.", "year": 2025, "url": "https://arxiv.org/abs/2506.15093", "path": "/sources/petrie-flexible-hardware-enabled-guarantees/" }, { "id": "S-1204", "title": "Technical Options for Flexible Hardware-Enabled Guarantees", "authors": "J. Petrie & O. Aarne", "year": 2025, "url": "https://arxiv.org/abs/2506.03409", "path": "/sources/petrie-technical-options-flexheg/" }, { "id": "S-1205", "title": "International Security Applications of Flexible Hardware-Enabled Guarantees", "authors": "O. Aarne & J. Petrie", "year": 2025, "url": "https://arxiv.org/abs/2506.15100", "path": "/sources/aarne-international-security-applications-flexheg/" }, { "id": "S-0057", "title": "Hardware-Enabled Governance Mechanisms: Developing Technical Solutions to Exempt Items Otherwise Classified Under Export Control Classification Numbers 3A090 and 4A090", "authors": "G. Kulp et al.", "year": 2024, "url": "https://www.rand.org/pubs/working_papers/WRA3056-1.html", "path": "/sources/kulp-hardware-enabled-governance-mechanisms/" }, { "id": "S-0056", "title": "Secure, Governable Chips: Using On-Chip Mechanisms to Manage National Security Risks from AI & Advanced Computing", "authors": "O. Aarne et al.", "year": 2024, "url": "https://www.cnas.org/publications/reports/secure-governable-chips", "path": "/sources/aarne-secure-governable-chips/" }, { "id": "S-0006", "title": "Hardware-Enabled Mechanisms for Verifying Responsible AI Development", "authors": "A. O'Gara et al.", "year": 2025, "url": "https://arxiv.org/abs/2505.03742", "path": "/sources/ogara-hardware-enabled-verifying-responsible-ai/" }, { "id": "S-1300", "title": "Fingerprinting All AI Cluster I/O Without Mutually Trusted Processors", "authors": "N. Cankaya et al.", "year": 2026, "url": "https://arxiv.org/abs/2606.10724", "path": "/sources/cankaya-fingerprinting-ai-cluster-io/" }, { "id": "S-0031", "title": "The Fundamentals and Feasibility of Secure Network Taps for Verifying AI Datacenter Use", "authors": "N. Cankaya", "year": 2026, "url": "https://nacicankaya.substack.com/p/research-note-the-fundamentals-and", "path": "/sources/cankaya-secure-network-taps/" }, { "id": "S-0018", "title": "A System Overview for Near-Term, Low-Trust AI Compute Verification", "authors": "N. Cankaya", "year": 2026, "url": "https://intelligence.org/wp-content/uploads/2026/06/A-system-overview-for-near-term-low-trust-AI-compute-verification.pdf", "path": "/sources/cankaya-system-overview-low-trust-compute-verification/" }, { "id": "S-0067", "title": "Verification Plan", "authors": "R. Dean", "year": 2026, "url": "https://ai-2040.com/supplements/verification-plan", "path": "/sources/dean-verification-plan/" }, { "id": "S-1312", "title": "Fitting a Network TAP to our Inference Verification Prototype", "authors": "Amodo Design", "year": 2026, "url": "https://amododesign.com/notes/2026-09-15-network-tap-inference-verification/", "path": "/sources/amodo-network-tap-inference-verification-prototype/" }, { "id": "S-1007", "title": "Amodo-Design/Inference-Recomputation-Prototype (GitHub repository)", "authors": "Amodo Design", "year": 2026, "url": "https://github.com/Amodo-Design/Inference-Recomputation-Prototype", "path": "/sources/amodo-inference-recomputation-prototype-code/" }, { "id": "S-1319", "title": "inference-verification: Inference Verification Prototype", "authors": "Singapore AI Safety Hub (SASH)", "year": 2026, "url": "https://github.com/sg-ai-safety-hub/inference-verification", "path": "/sources/sash-inference-verification-repo/" }, { "id": "S-1320", "title": "Internationalising AI Verification", "authors": "Singapore AI Safety Hub (SASH)", "year": 2026, "url": "https://www.aisafety.sg/research/internationalising-ai-verification", "path": "/sources/sash-internationalising-ai-verification/" }, { "id": "S-1310", "title": "Network Tapping for AI Verification: A Technical Assessment", "authors": "Amodo Design", "year": 2026, "url": "https://amododesign.com/notes/2026-05-03-network-tapping/", "path": "/sources/amodo-network-tapping-technical-assessment/" }, { "id": "S-1311", "title": "Network Traffic Hashing", "authors": "Amodo Design", "year": 2026, "url": "https://amododesign.com/notes/2026-07-03-network-traffic-hashing/", "path": "/sources/amodo-network-traffic-hashing/" } ] }