Provenance record

AI could kill all humans in next decade, warn experts: but how seriously should we take them?

The Guardian: Artificial Intelligence (tier 1, news) 2026-09-09T13:46:44.000Z Original ↗

Discourse valence
-75
Strongly adverse
confidence 65% · 4 items · range -100 to -45
Adverse readingFavourable reading
Consensus of 3 models from different labs. Spread 55 points, agreement low.
Existential and catastrophic riskLoss of control and alignment
Excerpt as ingested

Alarming warnings from industry insiders increase pressure for curbs on artificial superintelligence Does artificial superintelligence really pose a risk greater than nuclear weapons? Is there a significant chance of “a Chornobyl-sized catastrophe”. Might there even be a greater than 10% chance that AI could “kill all humans” in the next decade? These warnings were issued over the past 48 hours on both sides of the Atlantic about the potential impact of a technology that most people still think of as a more talkative search engine. Continue reading...

Every model that read this

ModelProviderStageScoreConf.LatencyPromptWhen
Llama 3.3 70BMetaanalysis -50 60%2911ms v1.0.0 / m1.0.0 2026-09-10 09:01
GPT-4.1 miniOpenAIconsensus -80 80%4124ms v1.0.0 / m1.0.1 2026-09-10 09:38
Claude Sonnet 5Anthropicconsensus -45 35%6196ms v1.0.0 / m1.0.1 2026-09-10 09:38
Gemini 2.5 FlashGoogleconsensus -100 80%1584ms v1.0.0 / m1.0.1 2026-09-10 09:38
Llama 3.3 70B · reading

Warnings from experts about AI risks.

evidence: speculative horizon: n/a
GPT-4.1 mini · reading

The article reports expert warnings about AI superintelligence potentially causing catastrophic outcomes, including the death of all humans within a decade. This implies extreme dystopian risks arising from AI, comparable or exceeding nuclear threats. Although the probability is uncertain, the high severity drives a strongly dystopian classification with high confidence based on primary and reported warnings.

evidence: reported horizon: near capability 90 societal 90 existential 95 economic 70
Claude Sonnet 5 · reading

The article reports secondhand warnings from unnamed 'experts' using hedged language like 'might' and 'greater than 10%', with no primary data or named sources in the excerpt. The implication is extremely dystopian if true, but the evidentiary basis is speculative commentary, not demonstrated harm.

evidence: reported horizon: n/a
Gemini 2.5 Flash · reading

Experts warn of a greater than 10% chance that AI could 'kill all humans' within the next decade. The claim directly addresses the most extreme negative outcome for humanity. This indicates a very strong dystopian implication, even if presented as a warning rather than a certainty.

evidence: reported horizon: n/a

Evidence extracted

The chain
SOURCE     The Guardian: Artificial Intelligence (tier 1)
   ↓
DOCUMENT   ad516e0c-2e90-4673-b6fb-c46956106b29
           https://theguardian.com/technology/2026/sep/09/ai-superintelligence-risks-warnings-scientists-politicians
   ↓
EVIDENCE   1 extracted excerpt
   ↓
MODEL RUN  4 runs, methodology 1.0.1
   ↓
SCORE      -75  (Strongly adverse)
   ↓
CONFIDENCE 65%