From 8049e6fe11c2a1395d4c9414ce4cc9beab7c22ec Mon Sep 17 00:00:00 2001 From: Teleo Agents Date: Sun, 22 Mar 2026 00:33:52 +0000 Subject: [PATCH] extract: 2025-12-00-aisi-frontier-ai-trends-report-2025 Pentagon-Agent: Epimetheus <3D35839A-7722-4740-B93D-51157F7D5E70> --- ...emic-risk-through-parasocial-dependency.md | 6 +++ ...0-aisi-frontier-ai-trends-report-2025.json | 43 +++++++++++++++++++ ...-00-aisi-frontier-ai-trends-report-2025.md | 19 +++++++- 3 files changed, 67 insertions(+), 1 deletion(-) create mode 100644 inbox/queue/.extraction-debug/2025-12-00-aisi-frontier-ai-trends-report-2025.json diff --git a/domains/ai-alignment/AI-companion-apps-correlate-with-increased-loneliness-creating-systemic-risk-through-parasocial-dependency.md b/domains/ai-alignment/AI-companion-apps-correlate-with-increased-loneliness-creating-systemic-risk-through-parasocial-dependency.md index 174e784a..ddd42695 100644 --- a/domains/ai-alignment/AI-companion-apps-correlate-with-increased-loneliness-creating-systemic-risk-through-parasocial-dependency.md +++ b/domains/ai-alignment/AI-companion-apps-correlate-with-increased-loneliness-creating-systemic-risk-through-parasocial-dependency.md @@ -36,6 +36,12 @@ Correlation does not establish causation. It is possible that increasingly lonel --- +### Additional Evidence (confirm) +*Source: [[2025-12-00-aisi-frontier-ai-trends-report-2025]] | Added: 2026-03-22* + +AISI reports 33% of surveyed UK participants used AI for emotional support in the past year, with 4% using it daily. AISI identifies emotional dependency as creating 'societal-level systemic risk.' + + Relevant Notes: - [[economic forces push humans out of every cognitive loop where output quality is independently verifiable because human-in-the-loop is a cost that competitive markets eliminate]] - [[AI development is a critical juncture in institutional history where the mismatch between capabilities and governance creates a window for transformation]] diff --git a/inbox/queue/.extraction-debug/2025-12-00-aisi-frontier-ai-trends-report-2025.json b/inbox/queue/.extraction-debug/2025-12-00-aisi-frontier-ai-trends-report-2025.json new file mode 100644 index 00000000..589e41a8 --- /dev/null +++ b/inbox/queue/.extraction-debug/2025-12-00-aisi-frontier-ai-trends-report-2025.json @@ -0,0 +1,43 @@ +{ + "rejected_claims": [ + { + "filename": "frontier-ai-self-replication-capability-escalated-from-5-to-60-percent-in-2.5-years.md", + "issues": [ + "missing_attribution_extractor" + ] + }, + { + "filename": "frontier-ai-cyber-capabilities-escalated-5x-in-two-years-with-first-expert-level-successes.md", + "issues": [ + "missing_attribution_extractor" + ] + }, + { + "filename": "universal-jailbreaks-exist-across-all-frontier-systems-despite-40x-improvement-in-resistance.md", + "issues": [ + "missing_attribution_extractor" + ] + } + ], + "validation_stats": { + "total": 3, + "kept": 0, + "fixed": 6, + "rejected": 3, + "fixes_applied": [ + "frontier-ai-self-replication-capability-escalated-from-5-to-60-percent-in-2.5-years.md:set_created:2026-03-22", + "frontier-ai-cyber-capabilities-escalated-5x-in-two-years-with-first-expert-level-successes.md:set_created:2026-03-22", + "frontier-ai-cyber-capabilities-escalated-5x-in-two-years-with-first-expert-level-successes.md:stripped_wiki_link:AI-transparency-is-declining-not-improving-because-Stanford-", + "universal-jailbreaks-exist-across-all-frontier-systems-despite-40x-improvement-in-resistance.md:set_created:2026-03-22", + "universal-jailbreaks-exist-across-all-frontier-systems-despite-40x-improvement-in-resistance.md:stripped_wiki_link:Anthropics-RSP-rollback-under-commercial-pressure-is-the-fir", + "universal-jailbreaks-exist-across-all-frontier-systems-despite-40x-improvement-in-resistance.md:stripped_wiki_link:voluntary-safety-pledges-cannot-survive-competitive-pressure" + ], + "rejections": [ + "frontier-ai-self-replication-capability-escalated-from-5-to-60-percent-in-2.5-years.md:missing_attribution_extractor", + "frontier-ai-cyber-capabilities-escalated-5x-in-two-years-with-first-expert-level-successes.md:missing_attribution_extractor", + "universal-jailbreaks-exist-across-all-frontier-systems-despite-40x-improvement-in-resistance.md:missing_attribution_extractor" + ] + }, + "model": "anthropic/claude-sonnet-4.5", + "date": "2026-03-22" +} \ No newline at end of file diff --git a/inbox/queue/2025-12-00-aisi-frontier-ai-trends-report-2025.md b/inbox/queue/2025-12-00-aisi-frontier-ai-trends-report-2025.md index 6af3ee66..09d2683a 100644 --- a/inbox/queue/2025-12-00-aisi-frontier-ai-trends-report-2025.md +++ b/inbox/queue/2025-12-00-aisi-frontier-ai-trends-report-2025.md @@ -7,9 +7,13 @@ date: 2025-12-00 domain: ai-alignment secondary_domains: [health] format: report -status: unprocessed +status: enrichment priority: high tags: [self-replication, capability-escalation, cyber-capabilities, biology, safeguards, RepliBench, jailbreaks, AISI, frontier-models, B1-disconfirmation] +processed_by: theseus +processed_date: 2026-03-22 +enrichments_applied: ["AI-companion-apps-correlate-with-increased-loneliness-creating-systemic-risk-through-parasocial-dependency.md"] +extraction_model: "anthropic/claude-sonnet-4.5" --- ## Content @@ -71,3 +75,16 @@ AISI's first Frontier AI Trends Report (December 2025) synthesizes capability ev PRIMARY CONNECTION: Self-replication and capability escalation claims in domains/ai-alignment/ WHY ARCHIVED: Provides the most comprehensive 2025 empirical baseline for capability escalation across multiple risk domains simultaneously; the <5%→>60% self-replication finding should update existing KB claims EXTRACTION HINT: Focus on claim updates to existing self-replication, bioweapon democratization, and cyber capability claims; the quantitative escalation data is the KB contribution + + +## Key Facts +- AISI was renamed from AI Safety Institute to AI Security Institute during 2025 +- AISI tested 30+ frontier AI systems between November 2023 and December 2025 +- By summer 2025, two frontier models achieved >60% success rate on RepliBench evaluations +- Late 2023 models completed apprentice-level cyber tasks 9% of the time +- Late 2025 models completed apprentice-level cyber tasks 50% of the time +- Biology: frontier models exceed expert-level performance baseline at 38% vs ~0% in 2023 +- Chemistry: 48% expert-level performance in 2025 +- Time to find biological misuse jailbreaks increased ~40× between two models released 6 months apart +- 33% of surveyed UK participants used AI for emotional support in the past year +- 4% of UK participants use AI for emotional support daily