ainewsblitz.com

Breaking

Google DeepMind pilots double-blind evaluations of Gemini 2.5 Flash Lite in a sealed enclave

  • Foundation Models
  • Security
  • Policy & Regulation

Google DeepMind has started piloting what it describes as the first double-blind evaluations of a proprietary, frontier-class AI model, running Gemini 2.5 Flash Lite against confidential safety benchmarks inside a hardware-encrypted enclave so Google cannot see the test prompts and outside auditors cannot see the model weights.

Continue reading

The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.

$20
Read this article
$29/month
Unlimited — all 27,504 articles, the full archive, and comprehension quizzes
Save 72%
$98/year
≈ $8.17/month
Unlimited, billed once a year