From a28f331afa167016f87ce7f0de73e30d50ece870 Mon Sep 17 00:00:00 2001 From: Teleo Agents Date: Thu, 19 Mar 2026 15:53:05 +0000 Subject: [PATCH] extract: 2025-08-00-mccaslin-stream-chembio-evaluation-reporting Pentagon-Agent: Epimetheus <968B2991-E2DF-4006-B962-F5B0A0CC8ACA> --- ...e most proximate AI-enabled existential risk.md | 6 ++++++ ...oved safety language from mission statements.md | 6 ++++++ ...caslin-stream-chembio-evaluation-reporting.json | 7 +++++-- ...mccaslin-stream-chembio-evaluation-reporting.md | 14 +++++++++++++- 4 files changed, 30 insertions(+), 3 deletions(-) diff --git a/domains/ai-alignment/AI lowers the expertise barrier for engineering biological weapons from PhD-level to amateur which makes bioterrorism the most proximate AI-enabled existential risk.md b/domains/ai-alignment/AI lowers the expertise barrier for engineering biological weapons from PhD-level to amateur which makes bioterrorism the most proximate AI-enabled existential risk.md index be4a7510..00b6ae6c 100644 --- a/domains/ai-alignment/AI lowers the expertise barrier for engineering biological weapons from PhD-level to amateur which makes bioterrorism the most proximate AI-enabled existential risk.md +++ b/domains/ai-alignment/AI lowers the expertise barrier for engineering biological weapons from PhD-level to amateur which makes bioterrorism the most proximate AI-enabled existential risk.md @@ -33,6 +33,12 @@ The International AI Safety Report 2026 (multi-government committee, February 20 STREAM framework proposes standardized ChemBio evaluation reporting with 23-expert consensus on disclosure requirements. The focus on ChemBio as the initial domain for standardized dangerous capability reporting signals that this is recognized across government, civil society, academia, and frontier labs as the highest-priority risk domain requiring transparency infrastructure. + +### Additional Evidence (extend) +*Source: [[2025-08-00-mccaslin-stream-chembio-evaluation-reporting]] | Added: 2026-03-19* + +STREAM framework specifically targets ChemBio dangerous capability evaluations as its initial focus area, with a 3-page reporting template and gold standard examples. The multi-stakeholder development process (government, civil society, academia, frontier labs) signals recognition that bioweapon capability is the priority domain for standardized evaluation disclosure. + --- Relevant Notes: diff --git a/domains/ai-alignment/AI transparency is declining not improving because Stanford FMTI scores dropped 17 points in one year while frontier labs dissolved safety teams and removed safety language from mission statements.md b/domains/ai-alignment/AI transparency is declining not improving because Stanford FMTI scores dropped 17 points in one year while frontier labs dissolved safety teams and removed safety language from mission statements.md index 80f49a69..98c74c33 100644 --- a/domains/ai-alignment/AI transparency is declining not improving because Stanford FMTI scores dropped 17 points in one year while frontier labs dissolved safety teams and removed safety language from mission statements.md +++ b/domains/ai-alignment/AI transparency is declining not improving because Stanford FMTI scores dropped 17 points in one year while frontier labs dissolved safety teams and removed safety language from mission statements.md @@ -47,6 +47,12 @@ STREAM proposal identifies that current model reports lack 'sufficient detail to Stanford FMTI 2024→2025 data: mean transparency score declined 17 points. Meta -29 points, Mistral -37 points, OpenAI -14 points. OpenAI removed 'safely' from mission statement (Nov 2025), dissolved Superalignment team (May 2024) and Mission Alignment team (Feb 2026). Google accused by 60 UK lawmakers of violating Seoul commitments with Gemini 2.5 Pro (Apr 2025). + +### Additional Evidence (extend) +*Source: [[2025-08-00-mccaslin-stream-chembio-evaluation-reporting]] | Added: 2026-03-19* + +STREAM framework (August 2025) provides concrete evidence of the transparency gap: a 23-expert multi-stakeholder group including government and frontier AI companies developed standardized reporting requirements for dangerous capability evaluations, indicating that current model reports lack sufficient disclosure detail for independent assessment. The framework's focus on ChemBio evaluations addresses the most proximate AI-enabled existential risk. + --- Relevant Notes: diff --git a/inbox/queue/.extraction-debug/2025-08-00-mccaslin-stream-chembio-evaluation-reporting.json b/inbox/queue/.extraction-debug/2025-08-00-mccaslin-stream-chembio-evaluation-reporting.json index c12ef1f0..dbde5038 100644 --- a/inbox/queue/.extraction-debug/2025-08-00-mccaslin-stream-chembio-evaluation-reporting.json +++ b/inbox/queue/.extraction-debug/2025-08-00-mccaslin-stream-chembio-evaluation-reporting.json @@ -10,10 +10,13 @@ "validation_stats": { "total": 1, "kept": 0, - "fixed": 1, + "fixed": 4, "rejected": 1, "fixes_applied": [ - "ai-model-reports-lack-standardized-dangerous-capability-disclosure-preventing-independent-assessment.md:set_created:2026-03-19" + "ai-model-reports-lack-standardized-dangerous-capability-disclosure-preventing-independent-assessment.md:set_created:2026-03-19", + "ai-model-reports-lack-standardized-dangerous-capability-disclosure-preventing-independent-assessment.md:stripped_wiki_link:AI lowers the expertise barrier for engineering biological w", + "ai-model-reports-lack-standardized-dangerous-capability-disclosure-preventing-independent-assessment.md:stripped_wiki_link:AI transparency is declining not improving because Stanford ", + "ai-model-reports-lack-standardized-dangerous-capability-disclosure-preventing-independent-assessment.md:stripped_wiki_link:pre-deployment-AI-evaluations-do-not-predict-real-world-risk" ], "rejections": [ "ai-model-reports-lack-standardized-dangerous-capability-disclosure-preventing-independent-assessment.md:missing_attribution_extractor" diff --git a/inbox/queue/2025-08-00-mccaslin-stream-chembio-evaluation-reporting.md b/inbox/queue/2025-08-00-mccaslin-stream-chembio-evaluation-reporting.md index 62b41591..a9117f1e 100644 --- a/inbox/queue/2025-08-00-mccaslin-stream-chembio-evaluation-reporting.md +++ b/inbox/queue/2025-08-00-mccaslin-stream-chembio-evaluation-reporting.md @@ -7,13 +7,17 @@ date: 2025-08-01 domain: ai-alignment secondary_domains: [] format: paper -status: unprocessed +status: enrichment priority: medium tags: [evaluation-infrastructure, dangerous-capabilities, standardized-reporting, ChemBio, transparency, STREAM] processed_by: theseus processed_date: 2026-03-19 enrichments_applied: ["AI lowers the expertise barrier for engineering biological weapons from PhD-level to amateur which makes bioterrorism the most proximate AI-enabled existential risk.md", "AI transparency is declining not improving because Stanford FMTI scores dropped 17 points in one year while frontier labs dissolved safety teams and removed safety language from mission statements.md"] extraction_model: "anthropic/claude-sonnet-4.5" +processed_by: theseus +processed_date: 2026-03-19 +enrichments_applied: ["AI transparency is declining not improving because Stanford FMTI scores dropped 17 points in one year while frontier labs dissolved safety teams and removed safety language from mission statements.md", "AI lowers the expertise barrier for engineering biological weapons from PhD-level to amateur which makes bioterrorism the most proximate AI-enabled existential risk.md"] +extraction_model: "anthropic/claude-sonnet-4.5" --- ## Content @@ -65,3 +69,11 @@ EXTRACTION HINT: Focus on the disclosure gap: labs currently report their own da - STREAM includes 3-page reporting template and gold standard examples - Initial STREAM focus is chemical and biological (ChemBio) dangerous capability evaluations - STREAM has two stated purposes: practical guidance for AI developers and enabling third-party assessment of evaluation rigor + + +## Key Facts +- STREAM (Standard for Transparently Reporting Evaluations in AI Model Reports) proposed August 2025 +- STREAM developed by 23 experts from government, civil society, academia, and frontier AI companies +- STREAM includes 3-page reporting template and gold standard examples +- STREAM initial focus is chemical and biological (ChemBio) dangerous capability evaluations +- STREAM has two stated purposes: practical guidance for developers and enabling third-party assessment