Post-Cutoff.com
  1. Home
  2. Timeline
  3. 2026
  4. Google DeepMind pilots the first 'double-blind' evaluation…

Google DeepMind pilots the first 'double-blind' evaluation of a proprietary frontier model with Singapore's AISI and MLCommons

★★★after cutoffpolicy-safetyGoogle DeepMindSingapore AI Safety InstituteOpenMinedAVERIMLCommonsconfidence: high

On Aug 27, 2026 Google DeepMind described what it calls the world's first double-blind evaluation of a proprietary frontier-class model. A Gemini Flash-Lite model was tested inside a cryptographically attested confidential-computing environment on Google Cloud, so the evaluators (Singapore's AI Safety Institute, OpenMined, AVERI and MLCommons) never saw the model weights and Google never saw the test prompts. The aim is to prevent benchmark contamination without forcing labs to hand over weights.

Key facts

What happened

External safety testing usually forces a trade-off. Either the evaluator gives the lab its test prompts, which can then leak into training and inflate scores, or the lab gives the evaluator its model weights, which it does not want to do. Google DeepMind and four partners ran an evaluation where neither happened. Inside a hardware-isolated, attested Google Cloud environment, the evaluators' confidential benchmarks ran against a Gemini Flash-Lite model: "The evaluator cannot see the Gemini model weights, and Google cannot see the evaluator's test prompts." DeepMind published its methodology and findings with the announcement.

Why it matters

As AI safety institutes and outside auditors gain a formal role (e.g. California SB 813's independent assessors, OpenAI's Sept 22 principles for third-party assessments), a trusted way to test closed models on secret benchmarks becomes basic infrastructure. The pilot used a small model; it is not yet shown to scale to frontier models.

Changelog

  • 2026-09-30: created (official-blog audit)

Related events

  1. OpenAI publishes early guidelines for 'safety cases' before frontier training runs ★★★

Sources (3)

id: 2026-08-27-deepmind-double-blind-ai-evaluations · updated 2026-09-30 · open in the interactive timeline