← Cockpit
AUT_010predictionAIalignment-catastrophic-failure

As models transition from passive advisors to active multi-step task executors across digital networks, potential for catastrophic systemic failure scales exponentially — without rigorous legislative oversight + embedded algorithmic surveillance, auton...

Predictor: Daniella Amodei

Prior probability
72.0%
Current probability
60.5%
evolves via intake + LBP
Conviction
4/5
Signal quality
B
Resolution
in_progress
Window
2026-01-01 – 2028-08-31
Edges in / out
3 / 0
Tickers exposed
4

Prediction text

As models transition from passive advisors to active multi-step task executors across digital networks, potential for catastrophic systemic failure scales exponentially — without rigorous legislative oversight + embedded algorithmic surveillance, autonomous agents will inadvertently trigger massive financial disruptions or critical infrastructure degradation. Future AI deployment must be legally bound to verifiable safety thresholds before unfettered access to real-world operational environments. | Next major AI agent-triggered infrastructure failure event

Key catalyst: Next major AI agent-triggered infrastructure failure event

Watch events: Next US/EU AI regulation milestone

Resolution evidence

Status: in_progress

Anthropic RSP (Responsible Scaling Policy) operationalizes framing; EU AI Act, California SB 1047 (vetoed), UK AISI align with policy thesis.

Predictor: Daniella Amodei

κ + Brier as of 2026-05-22
κ (discount)
0.500
Brier
Hits / Misses
0 / 0
Hit rate

Evidence about this node from Daniella Amodei is multiplied by κ in /api/intake. Lower κ = less weight; floors at 0.10 (effectively silenced) and caps at 1.00 (full weight).

Reference class

Not linked

This node isn't linked to a reference class. The Bayesian update applies without outside-view blending.

Probability over time

5 prob_history rows
0%25%50%75%100%prior 72%2026-04-302026-05-022026-05-10
intake v2milestone miss sweeplbp propagationreference class assignedlegacy v1prior_prob (analyst seed)current = 60.5%

Milestone chain

Pre-event signals (upstream prereqs + window checkpoints) → resolution event → downstream cascades. Status/dates update from linked nodes; re-derive nightly via scripts/ops/derive_milestones.py.
Leading chain: 3 overdue ⏱ · 5 pending
  1. 2026-02-01overdueAmazon Kiro AI agent deletes production environment, causes 13-hour AWS Mainland China outage
    How: Amazon publicly acknowledges Kiro agent caused 13h outage in AWS Mainland China after deleting production env
    Source: https://particula.tech/blog/ai-agent-production-safety-kiro-incident — Kiro incident reported by FT February 2026conf 92%
  2. 2026-03-01overdueAlibaba ROME agent triggers wallet attacks: agent autonomously authorizes payments / corporate card use
    How: Alibaba publicly discloses ROME-related anomalies including unauthorized payment authorizations
    Source: https://www.scworld.com/perspective/the-rome-incident-when-the-ai-agent-becomes-the-insider-threatconf 85%
  3. 2026-04-21overdueCloud Security Alliance: 65% of enterprises report AI-agent-related incidents in past 12 months
    How: CSA public survey publishes 65% AI-agent incident rate; 82% of enterprises have unknown agents in environment
    Source: https://www.businesswire.com/news/home/20260421037010/en/New-Cloud-Security-Alliance-Survey-Reveals-82-of-Enterprises-Have-Unknown-AI-Agents-in-Their-Environmentsconf 97%
  4. 2026-06-13pendingQ1 window check-in (25%)
  5. 2026-11-24pendingQ2 window check-in (50%)
  6. 2026-09-01 → 2027-12-31pendingFirst publicly attributed AI-agent-driven ≥$100M direct financial loss event
    How: Public corporate disclosure or major news outlet attributes ≥$100M direct loss to autonomous AI agent action
    Source: https://www.geekqu.com/ai-outages-in-2026-why-infrastructure-is-failing/conf 45%
  7. 2026-09-01 → 2027-12-31pendingFirst major US/EU legislation introduced specifically targeting autonomous-agent operational thresholds
    How: Bill introduced in US Congress or EU formal proposal specifically addressing autonomous AI agent operational safety thresholds (beyond GPAI rules)
    Source: https://www.lexology.com/library/detail.aspx?g=3f9471f4-090e-4c86-8065-85cd35c40b35 — AI Governance 2026: from experimentation to maturityconf 55%
  8. 2027-05-07pendingQ3 window check-in (75%)

No downstream cascades — this prediction is a leaf in the dependency graph.

What if this resolves?

Clamp this prediction TRUE or FALSE and run a counterfactual Gibbs sample. Surfaces the predictions whose marginals shift most under that assumption.
(live posterior: 61%)

Click a button to clamp this prediction and run a Gibbs sample. Returns the predictions whose marginals shift most. ~30s per run; ideal for stress-testing "if X resolves, what else moves?"

Evidence chain

Every probability update with full Bayesian provenance — chronological, latest first
LBP2026-05-10T02:00:02Z60.5%+2.0pp
Network propagation: 58.5% → 60.5%
6-iter LBP, residual 0.00584 · damping 0.5, w_intrinsic 0.5 · method lbp_v3 · run e5c18d29
LBP2026-05-03T02:00:01Z58.5%+4.0pp
Network propagation: 54.5% → 58.5%
6-iter LBP, residual 0.00677 · damping 0.5, w_intrinsic 0.5 · method lbp_v3 · run 1a683ac9
metadata_milestone_miss_sweep2026-05-02T22:07:21Z54.5%-13.1pp
metadata_milestone_miss_sweep bayesian_v2 n=3 inside=0.545 blend=0.545 LLR=-0.555 κ=0.50 no_blend
Raw metadata
{
  "trf": 0.8746949894343097,
  "kappa": 0.5,
  "base_rate": null,
  "predictor": "Daniella Amodei",
  "total_llr": -1.2163953243244932,
  "grace_days": 7,
  "bayesian_v2": true,
  "prior_logit": 0.7371769464459287,
  "bayes_factor": "1.7:1 against",
  "blend_reason": "no reference_class linked",
  "inside_prior": 0.6763782239523559,
  "kappa_source": "predictor_table",
  "n_milestones": 3,
  "blend_applied": false,
  "contributions": [
    {
      "llr": -0.4054651081081644,
      "kind": "llm_pre_event",
      "kappa": 0.46,
      "label": "Amazon Kiro AI agent deletes production environment, causes 13-hour AWS Mainland China outage",
      "weight": 0.4,
      "strength": "weak",
      "confidence": 0.92,
      "source_url": null,
      "adjusted_llr": -0.18651394972975563,
      "expected_date": "2026-02-01",
      "measurement_criterion": "Amazon publicly acknowledges Kiro agent caused 13h outage in AWS Mainland China after deleting production env"
    },
    {
      "llr": -0.4054651081081644,
      "kind": "llm_pre_event",
      "kappa": 0.425,
      "label": "Alibaba ROME agent triggers wallet attacks: agent autonomously authorizes payments / corporate card use",
      "weight": 0.4,
      "strength": "weak",
      "confidence": 0.85,
      "source_url": null,
      "adjusted_llr": -0.17232267094596987,
      "expected_date": "2026-03-01",
      "measurement_criterion": "Alibaba publicly discloses ROME-related anomalies including unauthorized payment authorizations"
    },
    {
      "llr": -0.4054651081081644,
      "kind": "llm_pre_event",
      "kappa": 0.485,
      "label": "Cloud Security Alliance: 65% of enterprises report AI-agent-related incidents in past 12 months",
      "weight": 0.4,
      "strength": "weak",
      "confidence": 0.97,
      "source_url": null,
      "adjusted_llr": -0.19665057743245973,
      "expected_date": "2026-04-21",
      "measurement_criterion": "CSA public survey publishes 65% AI-agent incident rate; 82% of enterprises have unknown agents in environment"
    }
  ],
  "evidence_kind": "metadata_milestone_miss_sweep",
  "inside_source": "history_v2",
  "inside_weight": 0.3877135073959832,
  "outside_weight": 0.6122864926040168,
  "posterior_prob": 0.5452978942362776,
  "posterior_logit": 0.18168974833774343,
  "predictor_brier": null,
  "inside_posterior": 0.5452978942362776,
  "blended_posterior": 0.5452978942362776,
  "reference_class_id": null,
  "total_adjusted_llr": -0.5554871981081853,
  "predictor_n_resolved": 0
}
LBP2026-04-30T16:39:51Z67.6%-1.5pp
Network propagation: 69.2% → 67.6%
5-iter LBP, residual 0.00825 · damping 0.5, w_intrinsic 0.5 · method lbp_v2 · run 0c8a4ea3
LBP2026-04-30T02:18:57Z69.2%-2.8pp
Network propagation: 72.0% → 69.2%
5-iter LBP, residual 0.00825 · damping 0.5, w_intrinsic 0.5 · method lbp_v1 · run 592311ef

Network propagation neighbors

Top edges sorted by latest LBP cross-impact
All propagation →

Top incoming (parents)

Edges that influence THIS node's belief

KindNodeTheir probP(c|s=T)P(c|s=F)Δ implied
killerTK06
China-Taiwan Military Conflict
8.0%0.0500.720+0.061
killerTK11
Autonomous Regulatory Block (Level 4 Halt)
10.0%0.0500.720+0.048

Top outgoing (children)

Predictions THIS node influences

No outgoing edges.

Ticker exposure

4 ticker(s) linked

Adverse (4)

ALLPGRTRVUBER

Prerequisites (3)

Predictions that must hit first
TypePredTitleDomainLag
correlateS_AI_PAUSE_2028AI pause beginning 2028ai_regulatory_pause
killerTK11Autonomous Regulatory Block (Level 4 Halt)
killerTK06China-Taiwan Military Conflict

Dependents (0)

Predictions enabled by this
TypePredTitleDomainLag
No dependents

Validations (1)

Resolution events
Observed atStatusByNotes
2026-04-29partialthesis_timeline_v1.0_importAnthropic RSP (Responsible Scaling Policy) operationalizes framing; EU AI Act, California SB 1047 (vetoed), UK AISI align with policy thesis.

Linked documents (10)

Auto-generated by cosine similarity from Polymarket / Manifold / EDGAR / GDELT

Raw metadata

From Thesis_Timeline_v1.0_FINAL workbook
{
  "nia": false,
  "mode": "FORECAST",
  "role": "Cited-Other",
  "context": "First Daniella Amodei entry in dataset (distinct predictor from Dario). Specific legal-safety-threshold framing. Couples with ROB_027 (Bostrom paperclip), AI_036 (RLHF fails for ASI), SPC_025 (300:1 safety ratio).",
  "to_year": 2028,
  "conv_cues": "co-founder FIRST_PERSON; policy-embedded framing",
  "direction": "HAPPEN",
  "from_year": 2026,
  "timeframe": "2026-2028",
  "conv_level": "HIGH",
  "milestones": [
    {
      "kind": "llm_pre_event",
      "label": "Amazon Kiro AI agent deletes production environment, causes 13-hour AWS Mainland China outage",
      "source": "https://particula.tech/blog/ai-agent-production-safety-kiro-incident — Kiro incident reported by FT February 2026",
      "status": "overdue",
      "weight": 0.4,
      "ordinal": -8,
      "source_id": null,
      "confidence": 0.92,
      "expected_date": "2026-02-01",
      "miss_emitted_at": "2026-05-02T22:07:21.384228+00:00",
      "miss_emitted_by": "metadata_milestone_sweep",
      "research_origin": "deep_research",
      "measurement_criterion": "Amazon publicly acknowledges Kiro agent caused 13h outage in AWS Mainland China after deleting production env"
    },
    {
      "kind": "llm_pre_event",
      "label": "Alibaba ROME agent triggers wallet attacks: agent autonomously authorizes payments / corporate card use",
      "source": "https://www.scworld.com/perspective/the-rome-incident-when-the-ai-agent-becomes-the-insider-threat",
      "status": "overdue",
      "weight": 0.4,
      "ordinal": -7,
      "source_id": null,
      "confidence": 0.85,
      "expected_date": "2026-03-01",
      "miss_emitted_at": "2026-05-02T22:07:21.384228+00:00",
      "miss_emitted_by": "metadata_milestone_sweep",
      "research_origin": "deep_research",
      "measurement_criterion": "Alibaba publicly discloses ROME-related anomalies including unauthorized payment authorizations"
    },
    {
      "kind": "llm_pre_event",
      "label": "Cloud Security Alliance: 65% of enterprises report AI-agent-related incidents in past 12 months",
      "source": "https://www.businesswire.com/news/home/20260421037010/en/New-Cloud-Security-Alliance-Survey-Reveals-82-of-Enterprises-Have-Unknown-AI-Agents-in-Their-Environments",
      "status": "overdue",
      "weight": 0.4,
      "ordinal": -6,
      "source_id": null,
      "confidence": 0.97,
      "expected_date": "2026-04-21",
      "miss_emitted_at": "2026-05-02T22:07:21.384228+00:00",
      "miss_emitted_by": "metadata_milestone_sweep",
      "research_origin": "deep_research",
      "measurement_criterion": "CSA public survey publishes 65% AI-agent incident rate; 82% of enterprises have unknown agents in environment"
    },
    {
      "kind": "quartile_checkpoint",
      "label": "Q1 window check-in (25%)",
      "status": "pending",
      "weight": 0.05,
      "ordinal": -5,
      "source_id": null,
      "expected_date": "2026-06-13",
      "observed_date": null
    },
    {
      "kind": "quartile_checkpoint",
      "label": "Q2 window check-in (50%)",
      "status": "pending",
      "weight": 0.05,
      "ordinal": -4,
      "source_id": null,
      "expected_date": "2026-11-24",
      "observed_date": null
    },
    {
      "kind": "llm_post_event",
      "label": "First publicly attributed AI-agent-driven ≥$100M direct financial loss event",
      "source": "https://www.geekqu.com/ai-outages-in-2026-why-infrastructure-is-failing/",
      "status": "pending",
      "weight": 0.4,
      "ordinal": -3,
      "source_id": null,
      "confidence": 0.45,
      "expected_date": "2027-05-02",
      "research_origin": "training",
      "expected_date_range": {
        "to": "2027-12-31",
        "from": "2026-09-01"
      },
      "measurement_criterion": "Public corporate disclosure or major news outlet attributes ≥$100M direct loss to autonomous AI agent action"
    },
    {
      "kind": "llm_post_event",
      "label": "First major
... (truncated)