Provenance record

The race to keep AI under control

Financial Times: Artificial Intelligence (tier 1, news) 2026-11-02T05:00:08.000Z Original ↗

Source note: Headlines and standfirsts only; full text is paywalled.

Discourse valence

-51

Adverse

confidence 65% · 5 items · range -65 to -35

-100

-50

+50

+100

-51

Adverse readingFavourable reading

Consensus of 5 models from different labs. Spread 30 points, agreement moderate.

Existential and catastrophic risk Governance and regulation Loss of control and alignment

Excerpt as ingested

Safety breaches are raising doubts about the ability of leading labs to keep increasingly powerful systems in check

Every model that read this

Model	Provider	Stage	Score	Conf.	Latency	Prompt	When
Llama 3.3 70B	Meta	analysis	-50	70%	2727ms	v1.0.0 / m1.0.1	2026-11-02 05:40
GPT-4.1 mini	OpenAI	consensus	-45	70%	1900ms	v1.0.0 / m1.0.1	2026-11-02 05:41
Claude Sonnet 5	Anthropic	consensus	-35	25%	3057ms	v1.0.0 / m1.0.1	2026-11-02 05:42
Gemini 2.5 Flash	Google	consensus	-60	80%	2054ms	v1.0.0 / m1.0.1	2026-11-02 05:42
Mistral Small 3.1 24B	Mistral AI	consensus	-65	80%	5943ms	v1.0.0 / m1.0.1	2026-11-02 05:42

Llama 3.3 70B · reading

Raises doubts about controlling powerful AI systems.

evidence: reported horizon: n/a

GPT-4.1 mini · reading

The claim highlights that safety breaches in AI labs are causing doubts about controlling powerful AI systems. This implies material risks to safety and governance but lacks details for a severe or existential score. Concerns about control and safety are significant but not fully evidenced here.

evidence: reported horizon: near capability 60 societal 0 existential 0 economic 0

Claude Sonnet 5 · reading

The headline and standfirst assert that safety breaches have occurred and cast doubt on labs' control capabilities, but no specifics on incidents, scale, or consequences are given. This implies a materially concerning trajectory, though thin sourcing limits confidence in magnitude.

evidence: speculative horizon: n/a

Gemini 2.5 Flash · reading

The item explicitly states that 'safety breaches are raising doubts' about the ability of leading AI labs to control increasingly powerful systems. This directly implies a significant risk to human flourishing due to uncontrolled AI.

evidence: reported horizon: n/a

Mistral Small 3.1 24B · reading

The headline suggests that AI systems are becoming more powerful and less controllable. This implies a loss of human control over AI.

evidence: speculative horizon: n/a

Evidence extracted

Safety breaches are raising doubts

The chain

SOURCE     Financial Times: Artificial Intelligence (tier 1)
   ↓
DOCUMENT   07dcf399-3622-46e0-8647-ed257dd7a9f9
           https://ig.ft.com/ai-safety
   ↓
EVIDENCE   1 extracted excerpt
   ↓
MODEL RUN  5 runs, methodology 1.0.1
   ↓
SCORE      -51  (Adverse)
   ↓
CONFIDENCE 65%