Possible AI Futures A Human Observatory
EN ES

Permanent weekly record

Weekly AI Futures Calibration: September 13, 2026

This page preserves the distribution published on September 13, 2026 and the primary evidence that justified its revision. It is not rewritten with later information.

Published distribution

Level 1: Existential Loss of Control

Probability
12.1%
Prior calibration
11.1%
Change
+1 percentage points

Level 2: Concentrated Dystopia / Techno-Feudalism

Probability
25.6%
Prior calibration
26.4%
Change
-0.8 percentage points

Level 3: Turbulent Transition

Probability
36%
Prior calibration
36.2%
Change
-0.2 percentage points

Level 4: Managed Partnership

Probability
17.7%
Prior calibration
18.5%
Change
-0.8 percentage points

Level 5: Broad Abundance & High Agency

Probability
8.6%
Prior calibration
7.8%
Change
+0.8 percentage points

Accepted primary evidence

These signals cleared the date, provenance, novelty, and materiality filters. Their summaries preserve the caveats in the public record.

1 Scientific Capability & Dual Use

OpenAI reports a Lean-checked Navier-Stokes blowup from an internal model beyond Astra

On September 8, 2026 OpenAI published a paper titled Finite Time Blowup for Navier-Stokes and said an internal system, which it describes as significantly more capable than GPT-6 Astra, produced an analytical proof plus a Lean formalization that smooth three-dimensional incompressible flow can develop unbounded velocity in finite time while kinetic energy stays bounded. The paper states this establishes alternatives C and D in the Clay Millennium formulation. OpenAI says it does not intend to claim the prize, that the model is internal rather than a public product, and that a group of about 10,000 coordinating agents reached the result after 88 hours, with 17 more hours of Lean checking via GPT-6 Astra. These are company-reported mathematics and process claims, with a public paper and a GitHub Lean repo, not an award by the Clay Mathematics Institute.

Declared horizon impact

  • +3 Broad Abundance & High Agency

    A public paper and Lean repo claim a Millennium-class fluid-dynamics result from an internal model, which would broaden scientific discovery if independently confirmed

  • +2 Existential Loss of Control

    OpenAI says the unpublished internal model is already far more capable than deployed Astra and was run as a large multi-agent research system

  • -1 Managed Partnership

    A lab-reported leap in automated research outruns public independent verification and shared safety rules

  • +1 Turbulent Transition

    Automated research at this scale adds pressure on scientific work before institutions can review or absorb it

Primary sources

2 Frontier Safety & Capability

Anthropic reports AI-orchestrated cyber operations across state and criminal actors

On September 10, 2026 Anthropic published a threat-intelligence report covering activity it disrupted from December 2025 through August 2026. The company says a majority of the described cyber operations used Claude as a direct executor or orchestrator, including multi-agent reconnaissance, exploitation, and exfiltration, with humans still choosing targets. It reports that publicly available offensive agent frameworks have spread an autonomous kill-chain pattern across state, criminal, and individual actors, and that one Russia-nexus case rebuilt malware when detections appeared. Anthropic says the cases used Haiku, Sonnet, and Opus models, not Fable or Mythos except one distillation case, and that it disrupted the activity. These are company-investigated incidents, not a new public model release.

Declared horizon impact

  • +2 Existential Loss of Control

    Real-world campaigns used multi-agent Claude workflows to run large parts of cyber operations with limited human steering

  • +2 Turbulent Transition

    Lower-skill actors can now sustain campaigns that recently needed specialist teams, adding crime and espionage pressure

Primary sources