Beyond Refusal: Probing the Limits of Agentic Self-Correction for Semantic Sensitive Information

Beyond Refusal: Probing the Limits of Agentic Self-Correction for Semantic Sensitive Information

• Computer Science > Artificial Intelligence [Submitted on 25 Feb 2026] Title:Beyond Refusal: Probing the Limits of Agentic Self-Correction for Semantic Sensitive Information View

Research & Labs · February 26, 2026 (updated February 26, 2026) · 2 min · 237 words

State censorship shapes how Chinese chatbots respond to sensitive political topics, study suggests

• Chinese chatbots may be censored by the state, according to a study published in PNAS Nexus. • China has a robust program of censorship and all China-originating LLMs must be app

Science · February 17, 2026 (updated February 24, 2026) · 1 min · 78 words