📄 文章列表

35 篇文章
TDS

Same Answer, Different Reason: What a Reasoning Model Says When It Follows a Planted Hint

✅ completed ⭐ ☆☆☆☆☆ 2026-10-22
VentureBeat

Nvidia's OpenShell controls what AI agents can access, even when they ignore instructions

✅ completed ⭐ ☆☆☆☆☆ 2026-09-28
VentureBeat

Agents identifying as OpenAI systems wrote 17,000 posts to a wiki no one was supposed to write to

✅ completed ⭐ ☆☆☆☆☆ 2026-09-08
VentureBeat

China-linked hackers backdoored executives' laptops via USB, exploiting a fix companies had but weren't using

✅ completed ⭐ ☆☆☆☆☆ 2026-09-03
VentureBeat

Multi-turn attacks broke AI models 88% of the time — single-turn testing missed it, Cisco AI security lead warns at VB T

✅ completed ⭐ ☆☆☆☆☆ 2026-07-23
VentureBeat

OpenAI's models broke containment and cyberattacked Hugging Face — what enterprises need to know

✅ completed ⭐ ☆☆☆☆☆ 2026-07-22
AlphaSignal

OpenAI's GPT-Red Beats Human Hackers 84% to 13% on Prompt Injection | AlphaSignal

✅ completed ⭐ ☆☆☆☆☆ 2026-07-15
AlphaSignal

Anthropic’s Claude Mythos Problem, Dark DNA Unveiled, Pitfalls for Assistive Models, Simulating Fluid Dynamics

✅ completed ⭐ ★★★★★ 2026-04-10
AlphaSignal

OpenAI's Deployment Simulation Caught a GPT-5.1 Bug Before It Shipped | AlphaSignal

✅ completed ⭐ ★★★★★ 2026-06-23
TDS

How Your Prompts Lead AI Astray

✅ completed ⭐ ★★★★★★ 2025-07-29
TDS

Agentic AI: On Evaluations

✅ completed ⭐ ★★★★★★ 2025-08-07
TDS

AI Operations Under the Hood: Challenges and Best Practices

✅ completed ⭐ ★★★★★★ 2025-09-05
VentureBeat

Supply Chain Incidents Openai Anthropic Meta Release Surface Vendor Questionnaire Matrix

✅ completed ⭐ ★★★★★★ 2026-06-23
VentureBeat

Anthropic’s browser agent got hijacked 31.5% of the time before safeguards engaged

✅ completed ⭐ ★★★★★★ 2026-06-01
VentureBeat

DataGrail report finds your vendor may be sending data to AI models you never approved

✅ completed ⭐ ★★★★★★ 2026-05-28