{"id":"3ece8de9133b5e17969a15c318dc41c16448df41d34b8ec2817b69757a616321","title":"METR proposes an example protocol for evaluating autonomous AI risks","argument":"METR published this example protocol on March 15, 2024, to evaluate AI models’ capability to pose autonomy-related risks. It focuses on systems that can complete harmful tasks end-to-end without humans as the bottleneck, including by procuring human assistance; it does not cover other risks such as harmful advice or deepfakes. METR proposes tasks of varying difficulty, model runs and scores that can be combined into a capability measure. The measure is intended to help decision-makers connect evaluation results with choices about model use, deployment, or further development, alongside threat modeling and risk tolerance. METR describes the protocol as an early warning, not a finished basis for setting mitigation thresholds: its authors say the science is at an early stage and do not yet offer a mapping they can defend. They suggest that crossing a threshold could call for pausing activity and developing better evaluations, while warning that changing tests after a threshold is crossed could weaken public accountability and invite gaming. The central unresolved issue is whether task performance and the proposed score can meaningfully track dangerous capabilities and real-world risk.","topicIds":["artificial-general-intelligence"],"domain":"other","claimType":"attributed reading","limits":"The protocol targets autonomous-system risks; METR says it would not catch other important risks, including those described in the source. METR says the scoring metric and aggregation scheme have not been empirically verified, and that human task duration is not a good measure of model difficulty or risk-relevant capability. Text as retrieved on October 25, 2026; originally published March 15, 2024.","speaker":{"name":"METR","roleAtSource":"Publisher"},"source":{"title":"Example autonomy evaluation protocol","publisher":"METR","format":"Article","url":"https://metr.org/blog/2024-03-15-example-autonomy-evaluation-protocol/","publishedAt":"2024-03-15T11:00:00.000Z","dateBasis":"Date supplied by the publisher’s feed."},"evidence":[{"text":"METR says the protocol targets risks from autonomous AI systems that can carry out substantive harmful tasks without humans as the bottleneck.","excerpt":"systems that are able to accomplish substantive harmful tasks end-to-end on their own without humans being the bottleneck","url":"https://metr.org/blog/2024-03-15-example-autonomy-evaluation-protocol/","relationship":"Source-stated"},{"text":"METR says the protocol does not address several other kinds of AI risk and calls for separate evaluations to address them.","excerpt":"These types of risks are important, and this protocol would not catch them. We think separate evaluations should be conducted to ensure that these risks are also addressed.","url":"https://metr.org/blog/2024-03-15-example-autonomy-evaluation-protocol/","relationship":"Source-stated"},{"text":"METR says the protocol is an early warning and that its authors cannot yet defend an overall mapping from scores to mitigations.","excerpt":"The science of evaluation for autonomous capabilities is still in very early stages, so we don’t yet have overall suggestions for what this mapping should be that we feel able to defend.","url":"https://metr.org/blog/2024-03-15-example-autonomy-evaluation-protocol/","relationship":"Source-stated"},{"text":"METR warns that changing evaluations after a threshold is crossed may weaken public accountability and create opportunities to game the updated tests.","excerpt":"AI developers might face intense pressures to (and have ample opportunity to) construct the new evals to produce the desired result, or to game them.","url":"https://metr.org/blog/2024-03-15-example-autonomy-evaluation-protocol/","relationship":"Source-stated"}],"sourceMedia":{"status":"unchecked","sourceUrl":"https://metr.org/blog/2024-03-15-example-autonomy-evaluation-protocol/","checkedAt":null,"items":[]},"assumptions":[],"openQuestion":{"attribution":"Open question","text":"How well would performance on this task suite predict the capabilities and risks of autonomous AI systems outside the evaluations?"},"whatWouldChangeView":{"attribution":"Change condition not recorded","text":""},"editorialComparisons":[],"live":{"title":"METR proposes an example protocol for evaluating autonomous AI risks","summary":"METR published this example protocol on March 15, 2024, to evaluate AI models’ capability to pose autonomy-related risks. It focuses on systems that can complete harmful tasks end-to-end without humans as the bottleneck, including by procuring human assistance; it does not cover other risks such as harmful advice or deepfakes. METR proposes tasks of varying difficulty, model runs and scores that can be combined into a capability measure. The measure is intended to help decision-makers connect evaluation results with choices about model use, deployment, or further development, alongside threat modeling and risk tolerance. METR describes the protocol as an early warning, not a finished basis for setting mitigation thresholds: its authors say the science is at an early stage and do not yet offer a mapping they can defend. They suggest that crossing a threshold could call for pausing activity and developing better evaluations, while warning that changing tests after a threshold is crossed could weaken public accountability and invite gaming. The central unresolved issue is whether task performance and the proposed score can meaningfully track dangerous capabilities and real-world risk.","question":"How well would performance on this task suite predict the capabilities and risks of autonomous AI systems outside the evaluations?","claims":[{"text":"METR says the protocol targets risks from autonomous AI systems that can carry out substantive harmful tasks without humans as the bottleneck.","excerpt":"systems that are able to accomplish substantive harmful tasks end-to-end on their own without humans being the bottleneck","support":"source-stated"},{"text":"METR says the protocol does not address several other kinds of AI risk and calls for separate evaluations to address them.","excerpt":"These types of risks are important, and this protocol would not catch them. We think separate evaluations should be conducted to ensure that these risks are also addressed.","support":"source-stated"},{"text":"METR says the protocol is an early warning and that its authors cannot yet defend an overall mapping from scores to mitigations.","excerpt":"The science of evaluation for autonomous capabilities is still in very early stages, so we don’t yet have overall suggestions for what this mapping should be that we feel able to defend.","support":"source-stated"},{"text":"METR warns that changing evaluations after a threshold is crossed may weaken public accountability and create opportunities to game the updated tests.","excerpt":"AI developers might face intense pressures to (and have ample opportunity to) construct the new evals to produce the desired result, or to game them.","support":"source-stated"}],"limitations":["The protocol targets autonomous-system risks; METR says it would not catch other important risks, including those described in the source.","METR says the scoring metric and aggregation scheme have not been empirically verified, and that human task duration is not a good measure of model difficulty or risk-relevant capability.","Text as retrieved on October 25, 2026; originally published March 15, 2024."],"mediaId":null,"correctionReason":null,"id":"3ece8de9133b5e17969a15c318dc41c16448df41d34b8ec2817b69757a616321","publicationId":"48a4a042-db60-9e20-4f1e-415553f8a6a8","version":1,"sourceUrl":"https://metr.org/blog/2024-03-15-example-autonomy-evaluation-protocol/","sourceTitle":"Example autonomy evaluation protocol","publisher":"METR","publishedAt":"2024-03-15T11:00:00.000Z","originalPublished":{"value":"2024-03-15T11:00:00.000Z","precision":"instant","basis":"backfill:feed_published"},"firstObservedAt":"2026-10-18T03:53:24.687Z","createdAt":"2026-11-20T23:03:10.108Z","historical":true,"entityIds":["metr"],"topicIds":["artificial-general-intelligence"],"media":[],"sourceMedia":{"status":"unchecked","sourceUrl":"https://metr.org/blog/2024-03-15-example-autonomy-evaluation-protocol/","checkedAt":null,"items":[]},"editorialDecision":{"outcome":"accept","mediaId":null,"inputHash":"97e39aafb7e7325cd3485f1d87b7cf95cdaf81bb7f59ff7be16a0a6c0f5bb448","rubricVersion":"unrecorded","provider":"vercel","model":"typesafe-ai/jev","reasonCodes":["bridge_admitted_by_model"],"confidence":0.5,"decidedAt":"2026-10-25T03:07:36.494Z"},"provenance":{"textHash":"39fa4a36a8aa9731adb39492d8d66bcfd2c3da8f79a32256e0a6a85ba3a53936","observationId":"c719b81b1ece81a7bef3fd159e8448e193c1b6d83ac32fe1323caf91f83ab050","retrievedAt":"2026-10-25T03:09:42.887Z","kind":"publisher-page","preparer":{"model":"openai/gpt-6-luna","runId":"1d9ba848-a7a5-38b0-6c97-c562d8873e67"},"reviewer":{"model":"anthropic/claude-opus-5.5","runId":"7e9cdf9f-2333-7408-b584-87c0bdb72295"}},"audit":{"status":"passed","reviewedAt":"2026-11-20T22:58:26.165Z","model":"anthropic/claude-opus-5.5"}},"canonicalUrl":"https://escapemeridian.com/readings/metr-proposes-an-example-protocol-for-evaluating-autonomous-ai-3ece8de9","formats":{"html":"/readings/metr-proposes-an-example-protocol-for-evaluating-autonomous-ai-3ece8de9","markdown":"/readings/metr-proposes-an-example-protocol-for-evaluating-autonomous-ai-3ece8de9.md","json":"/readings/metr-proposes-an-example-protocol-for-evaluating-autonomous-ai-3ece8de9.json"},"attribution":"Headline and summary by Escape Meridian, prepared by Luna from the attributed source. A source check is not independent confirmation of the source’s claims.","maintainedGuideLinks":[],"conditions":[],"readerContext":{"version":"reader-context-v1","attribution":"Reference context selected by Escape Meridian; separate from source evidence. Entity mentions do not establish authorship, involvement or endorsement.","terms":[],"entities":[{"id":"metr","name":"METR","kind":"organization","context":"METR studies what AI systems can do and how reliably they complete tasks, including work with human developers.","referenceUrls":["https://metr.org/"],"path":"/entities/metr","json":"/entities/metr.json","markdown":"/entities/metr.md"}]},"researchQuestions":[],"withdrawn":false,"references":{"schemaVersion":1,"target":{"type":"reading","id":"3ece8de9133b5e17969a15c318dc41c16448df41d34b8ec2817b69757a616321","revision":"48a4a042-db60-9e20-4f1e-415553f8a6a8","revisionKind":"publication-id","title":"METR proposes an example protocol for evaluating autonomous AI risks","url":"https://escapemeridian.com/readings/metr-proposes-an-example-protocol-for-evaluating-autonomous-ai-3ece8de9"},"source":{"type":"source","id":"https://metr.org/blog/2024-03-15-example-autonomy-evaluation-protocol/","revision":"39fa4a36a8aa9731adb39492d8d66bcfd2c3da8f79a32256e0a6a85ba3a53936","revisionKind":"sha256-source-text","title":"Example autonomy evaluation protocol","url":"https://metr.org/blog/2024-03-15-example-autonomy-evaluation-protocol/"},"status":"available","items":[],"next":null,"scope":"Explicit citations in public readings to this reading. Shared topics are not citations. Other record types need an explicit reference adapter."},"revisions":[{"id":"3ece8de9133b5e17969a15c318dc41c16448df41d34b8ec2817b69757a616321","title":"METR proposes an example protocol for evaluating autonomous AI risks","argument":"METR published this example protocol on March 15, 2024, to evaluate AI models’ capability to pose autonomy-related risks. It focuses on systems that can complete harmful tasks end-to-end without humans as the bottleneck, including by procuring human assistance; it does not cover other risks such as harmful advice or deepfakes. METR proposes tasks of varying difficulty, model runs and scores that can be combined into a capability measure. The measure is intended to help decision-makers connect evaluation results with choices about model use, deployment, or further development, alongside threat modeling and risk tolerance. METR describes the protocol as an early warning, not a finished basis for setting mitigation thresholds: its authors say the science is at an early stage and do not yet offer a mapping they can defend. They suggest that crossing a threshold could call for pausing activity and developing better evaluations, while warning that changing tests after a threshold is crossed could weaken public accountability and invite gaming. The central unresolved issue is whether task performance and the proposed score can meaningfully track dangerous capabilities and real-world risk.","topicIds":["artificial-general-intelligence"],"domain":"other","claimType":"attributed reading","limits":"The protocol targets autonomous-system risks; METR says it would not catch other important risks, including those described in the source. METR says the scoring metric and aggregation scheme have not been empirically verified, and that human task duration is not a good measure of model difficulty or risk-relevant capability. Text as retrieved on October 25, 2026; originally published March 15, 2024.","speaker":{"name":"METR","roleAtSource":"Publisher"},"source":{"title":"Example autonomy evaluation protocol","publisher":"METR","format":"Article","url":"https://metr.org/blog/2024-03-15-example-autonomy-evaluation-protocol/","publishedAt":"2024-03-15T11:00:00.000Z","dateBasis":"Date supplied by the publisher’s feed."},"evidence":[{"text":"METR says the protocol targets risks from autonomous AI systems that can carry out substantive harmful tasks without humans as the bottleneck.","excerpt":"systems that are able to accomplish substantive harmful tasks end-to-end on their own without humans being the bottleneck","url":"https://metr.org/blog/2024-03-15-example-autonomy-evaluation-protocol/","relationship":"Source-stated"},{"text":"METR says the protocol does not address several other kinds of AI risk and calls for separate evaluations to address them.","excerpt":"These types of risks are important, and this protocol would not catch them. We think separate evaluations should be conducted to ensure that these risks are also addressed.","url":"https://metr.org/blog/2024-03-15-example-autonomy-evaluation-protocol/","relationship":"Source-stated"},{"text":"METR says the protocol is an early warning and that its authors cannot yet defend an overall mapping from scores to mitigations.","excerpt":"The science of evaluation for autonomous capabilities is still in very early stages, so we don’t yet have overall suggestions for what this mapping should be that we feel able to defend.","url":"https://metr.org/blog/2024-03-15-example-autonomy-evaluation-protocol/","relationship":"Source-stated"},{"text":"METR warns that changing evaluations after a threshold is crossed may weaken public accountability and create opportunities to game the updated tests.","excerpt":"AI developers might face intense pressures to (and have ample opportunity to) construct the new evals to produce the desired result, or to game them.","url":"https://metr.org/blog/2024-03-15-example-autonomy-evaluation-protocol/","relationship":"Source-stated"}],"sourceMedia":{"status":"unchecked","sourceUrl":"https://metr.org/blog/2024-03-15-example-autonomy-evaluation-protocol/","checkedAt":null,"items":[]},"assumptions":[],"openQuestion":{"attribution":"Open question","text":"How well would performance on this task suite predict the capabilities and risks of autonomous AI systems outside the evaluations?"},"whatWouldChangeView":{"attribution":"Change condition not recorded","text":""},"editorialComparisons":[],"live":{"title":"METR proposes an example protocol for evaluating autonomous AI risks","summary":"METR published this example protocol on March 15, 2024, to evaluate AI models’ capability to pose autonomy-related risks. It focuses on systems that can complete harmful tasks end-to-end without humans as the bottleneck, including by procuring human assistance; it does not cover other risks such as harmful advice or deepfakes. METR proposes tasks of varying difficulty, model runs and scores that can be combined into a capability measure. The measure is intended to help decision-makers connect evaluation results with choices about model use, deployment, or further development, alongside threat modeling and risk tolerance. METR describes the protocol as an early warning, not a finished basis for setting mitigation thresholds: its authors say the science is at an early stage and do not yet offer a mapping they can defend. They suggest that crossing a threshold could call for pausing activity and developing better evaluations, while warning that changing tests after a threshold is crossed could weaken public accountability and invite gaming. The central unresolved issue is whether task performance and the proposed score can meaningfully track dangerous capabilities and real-world risk.","question":"How well would performance on this task suite predict the capabilities and risks of autonomous AI systems outside the evaluations?","claims":[{"text":"METR says the protocol targets risks from autonomous AI systems that can carry out substantive harmful tasks without humans as the bottleneck.","excerpt":"systems that are able to accomplish substantive harmful tasks end-to-end on their own without humans being the bottleneck","support":"source-stated"},{"text":"METR says the protocol does not address several other kinds of AI risk and calls for separate evaluations to address them.","excerpt":"These types of risks are important, and this protocol would not catch them. We think separate evaluations should be conducted to ensure that these risks are also addressed.","support":"source-stated"},{"text":"METR says the protocol is an early warning and that its authors cannot yet defend an overall mapping from scores to mitigations.","excerpt":"The science of evaluation for autonomous capabilities is still in very early stages, so we don’t yet have overall suggestions for what this mapping should be that we feel able to defend.","support":"source-stated"},{"text":"METR warns that changing evaluations after a threshold is crossed may weaken public accountability and create opportunities to game the updated tests.","excerpt":"AI developers might face intense pressures to (and have ample opportunity to) construct the new evals to produce the desired result, or to game them.","support":"source-stated"}],"limitations":["The protocol targets autonomous-system risks; METR says it would not catch other important risks, including those described in the source.","METR says the scoring metric and aggregation scheme have not been empirically verified, and that human task duration is not a good measure of model difficulty or risk-relevant capability.","Text as retrieved on October 25, 2026; originally published March 15, 2024."],"mediaId":null,"correctionReason":null,"id":"3ece8de9133b5e17969a15c318dc41c16448df41d34b8ec2817b69757a616321","publicationId":"48a4a042-db60-9e20-4f1e-415553f8a6a8","version":1,"sourceUrl":"https://metr.org/blog/2024-03-15-example-autonomy-evaluation-protocol/","sourceTitle":"Example autonomy evaluation protocol","publisher":"METR","publishedAt":"2024-03-15T11:00:00.000Z","originalPublished":{"value":"2024-03-15T11:00:00.000Z","precision":"instant","basis":"backfill:feed_published"},"firstObservedAt":"2026-10-18T03:53:24.687Z","createdAt":"2026-11-20T23:03:10.108Z","historical":true,"entityIds":["metr"],"topicIds":["artificial-general-intelligence"],"media":[],"sourceMedia":{"status":"unchecked","sourceUrl":"https://metr.org/blog/2024-03-15-example-autonomy-evaluation-protocol/","checkedAt":null,"items":[]},"editorialDecision":{"outcome":"accept","mediaId":null,"inputHash":"97e39aafb7e7325cd3485f1d87b7cf95cdaf81bb7f59ff7be16a0a6c0f5bb448","rubricVersion":"unrecorded","provider":"vercel","model":"typesafe-ai/jev","reasonCodes":["bridge_admitted_by_model"],"confidence":0.5,"decidedAt":"2026-10-25T03:07:36.494Z"},"provenance":{"textHash":"39fa4a36a8aa9731adb39492d8d66bcfd2c3da8f79a32256e0a6a85ba3a53936","observationId":"c719b81b1ece81a7bef3fd159e8448e193c1b6d83ac32fe1323caf91f83ab050","retrievedAt":"2026-10-25T03:09:42.887Z","kind":"publisher-page","preparer":{"model":"openai/gpt-6-luna","runId":"1d9ba848-a7a5-38b0-6c97-c562d8873e67"},"reviewer":{"model":"anthropic/claude-opus-5.5","runId":"7e9cdf9f-2333-7408-b584-87c0bdb72295"}},"audit":{"status":"passed","reviewedAt":"2026-11-20T22:58:26.165Z","model":"anthropic/claude-opus-5.5"}}}],"audits":[{"publicationId":"48a4a042-db60-9e20-4f1e-415553f8a6a8","verdict":"passed","findings":[{"dimension":"attribution","status":"pass","explanation":"The reading consistently attributes the protocol, its scope, its caveats and its accountability warning to METR, the publisher of this single blog document. It does not present METR's claims as established fact.","excerpt":"This is an example protocol for evaluating an AI model’s capability to pose autonomy-related risks."},{"dimension":"summary","status":"pass","explanation":"The reading leads with the actor and the development. It explains autonomy risk in plain terms, describes the task, run and score structure, and states the intended decision use. It preserves METR's self-described early-warning status, its admission that it has no defensible mapping, and its tension between pausing and accountability. The question follows from METR's own caveats about the metric. It could also have mentioned the conservative bound of under 1 hour, but that omission does not distort the reading.","excerpt":"This evaluation protocol could be rapidly improved, and is intended only as a conservative early warning"},{"dimension":"date","status":"pass","explanation":"originalPublished is 2024-03-15T11:00Z, and the reading states March 15, 2024, which matches the UTC day and the URL path. Nothing in the reading uses hindsight or later events. The source's plans, such as future updates, are not presented as outcomes. The evidence is a publisher page, not an archive copy, so no archive disclosure is required. The retrieval-date note is accurate.","excerpt":null},{"dimension":"support","status":"pass","explanation":"Each claim is backed by a verbatim excerpt. The limitations are grounded in the source, including the unverified metric and the human-hours caveat. The phrase 'not a good measure' closely tracks the source's 'not a great metric'. The pausing recommendation is hedged, matching the source's 'often'. The reading adds no unsupported benefits or forecasts.","excerpt":"The metric and aggregation scheme we use hasn’t been empirically verified"},{"dimension":"media","status":"pass","explanation":"No media is selected, and no verified candidates were supplied. The source page carries no media items, so omitting media is appropriate.","excerpt":null}],"createdAt":"2026-11-20T22:58:26.165Z","model":"anthropic/claude-opus-5.5"}],"correctionUrl":"https://escapemeridian.com/corrections/"}