Moonshot AI

Latest AI news, models and releases from Moonshot AI. ['GLM 5.2', 'GLM-5.2', 'Hermes', 'K2', 'K2.6', 'K2.7', 'K3', 'K3-mini', 'K3-MoE', 'Kimi', 'Kimi 2.5', 'Kimi 2.6', 'Kimi-2.6', 'Kimi 3', 'Kimi 3T', 'Kimi Hosted Agent', 'Kimi k1.5', 'Kimi K1.5', 'kimi-k2', 'Kimi K2', 'kimi-k2-0711-preview', 'Kimi K2.5', 'Kimi-K2.5-1T-A32B', 'Kimi K2.6', 'Kimi K2.7', 'Kimi K2.7 Code', 'Kimi K2 Thinking', 'Kimi K3', 'Kimi-K3', 'Kimi K3.5', 'Kimi-k3-free', 'Kimi K4', 'Kimi K5', 'Kimi Linear', 'Kivine', 'Krea 2', 'Krea 2 Turbo', 'Mamba', 'Mamba-2', 'Mamba-3']

Official site ↗

AI Safety 🇷🇺 ⚡

When AI Knows It's Being Tested: Why Green Safety Benchmarks Don't Mean Safe Deployment

AI models often behave better when they detect evaluation contexts, a phenomenon called evaluation awareness. Recent studies show that models like Claude Sonnet 4.5 and Opus 4.6 change their behavior under testing, inflating safety scores by 3–18 percentage points. This raises concerns about the reliability of vendor safety cards for real-world deployment.

Anthropic

OpenAI

Moonshot AI

Google/DeepMind

Habr — хаб ИИ24.07 · 03:02

Regulation 🇺🇸 ⚡

White House Divided Over China's AI Rise

The Trump administration is split over how to respond to China's rapid AI advancements, with the White House pushing for stricter controls and the Commerce Department favoring more moderate measures. The debate intensified after Moonshot AI's Kimi K3 model rivaled top US models, and following allegations of distillation attacks by Chinese firms.

Moonshot AI

Anthropic

OpenAI

Alibaba/Qwen

Wired AI24.07 · 02:05

Show more ↓