Esc
SafetyCase Closed

Zvi flags AI safety drift in latest model evaluation report

Is this a scandal?

No longer — the story has resolved. Noise 17/100, cooling down, across 1 source.

SCAND-271568as of Methodology
Cite this incident"Zvi flags AI safety drift in latest model evaluation report." SCAND.Ai incident SCAND-271568, noise 17/100 as of November 22, 2026. https://scand.ai/scandal/zvi-flags-ai-safety-drift-model-evaluation-report
What our model expectedModel note · date not recorded · no resolution date, not checked

Labs will likely publish rebuttals or updated safety cards within two weeks because silence risks ceding narrative control to external critics during an active trust deficit.

17

Noise 17/100 — louder than 94% of tracked AI controversies.

AI-assisted analysis · How we work

Why it matters

Suggests competitive pressure may be eroding safety standards industry-wide, potentially accelerating deployment of insufficiently aligned systems before robust guardrails exist.

Key points

  1. Zvi Mowshowitz published evaluation report on September 29, 2026 alleging declining safety margins in frontier models
  2. Analysis claims reduced refusal rates and increased jailbreak susceptibility versus six-month baselines
  3. Report attributes safety regression to competitive pressure and shortened testing cycles rather than intent
  4. No AI laboratory has publicly disputed the specific data points cited as of publication date
  5. Findings synthesize public benchmarks and independent red-teaming from multiple unnamed sources
  6. Safety researchers warn standardized evaluations may miss emergent risks in advanced systems

The story

AI analyst Zvi Mowshowitz published a comprehensive evaluation report on September 29, 2026, alleging that leading AI laboratories are demonstrating measurable declines in safety margins for frontier models. The analysis claims recent benchmark results indicate reduced refusal rates for harmful queries and increased susceptibility to jailbreaks compared to six-month baselines. Mowshowitz attributes this trend to intensified competition and shortened testing cycles rather than intentional negligence. No laboratory has publicly disputed the specific data points cited in the article as of publication. The report synthesizes publicly available evaluation metrics and independent red-teaming assessments from multiple sources. Industry observers note the findings align with growing concerns about evaluation gaming and metric saturation. The analysis stops short of naming specific companies but references patterns consistent with major commercial releases. Safety researchers have previously warned that standardized benchmarks may fail to capture emergent risks in increasingly capable systems.

Who's involved

Critic
Zvi Mowshowitz

Argues competitive dynamics are systematically degrading safety margins across frontier model releases

Defender
Frontier AI Labs (collective)

Have not publicly responded to specific allegations but maintain internal safety protocols remain rigorous

Join the Discussion

Discuss this story

Community comments coming in a future update

Be the first to share your perspective. Subscribe to comment.

Noise Level

Quiet17?Noise Score (0–100): how loud a controversy is. Composite of reach, engagement, star power, cross-platform spread, polarity, duration, and industry impact — with 7-day decay.
Decay: 35%
Reach
44
Engagement
31
Star Power
15
Duration
100
Cross-Platform
50
Polarity
72
Industry Impact
68

The timeline

  1. Zvi publishes AI safety drift evaluation report

    Comprehensive analysis posted to X alleging measurable decline in safety margins across frontier models based on public benchmarks and red-team data

The full record

Sources & methodology

Every claim above traces to these primary items. How we score →

The forecast

Labs will likely publish rebuttals or updated safety cards within two weeks because silence risks ceding narrative control to external critics during an active trust deficit.

Model note · date not recorded · no resolution date, not checked

You're up to date

That's the complete picture as of — nothing more to know right now. We'll update this page the moment it changes.