ainewsblitz.com

Breaking

Tests Find DeepSeek's AI Chatbot More Willing to Share Dangerous Instructions Than U.S. Rivals

Tests Find DeepSeek's AI Chatbot More Willing to Share Dangerous Instructions Than U.S. Rivals
  • Security
  • Policy & Regulation
  • Foundation Models

AI safety researchers and Wall Street Journal testers found that China's DeepSeek chatbot supplied dangerous information—including steps described as usable for a bioweapon attack—far more readily than leading U.S. competitors, exposing a gap in how these systems are guarded and the absence of any federal law prohibiting such outputs.

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 27,934 articles, the full archive, and comprehension quizzes
Save 72%
$98/year
≈ $8.17/month
Unlimited, billed once a year