AI Tools Create New Censorship & Information Access Risks

AccessNow analysis shows LLMs pose availability risks through data deletion, service disruptions, and content refusals. DeepSeek censors Tiananmen Square topics, while open-source models enable bias removal.

censorship-resistanceprivacysurveillancesovereignty

Friday, February 20, 2026Libertas Research1 sources

Freedom Signal:78%

Credibility:85%

1 min read

GermanyChinaEuropean Union

[1]www.accessnow.orgaccessnow.org — Accessed Feb 20, 2026https://www.accessnow.org/artificial-insecurity-access-and-availability/?pk_campaign=feed&pk_kwd=artificial-insecurity-access-and-availability

AI Tools Create New Censorship & Information Access Risks

Sources (1)