robuster.ai

Sources

Every claim, checkable.

A site about keeping AI honest should show its work. These are the studies, rulings, and reports behind the claims on this site.

  1. Kalai, Nachum, Vempala, Zhang. "Why Language Models Hallucinate." OpenAI, September 2025. openai.com↩ Back to Hallucination
  2. Dziri et al. "Faith and Fate: Limits of Transformers on Compositionality." NeurIPS 2023. arxiv.org/abs/2305.18654↩ Back to Multi-step inference
  3. Liu et al. "Lost in the Middle: How Language Models Use Long Contexts." TACL 2023. arxiv.org/abs/2307.03172↩ Back to Drift
  4. Hong, Troynikov, Huber. "Context Rot: How Increasing Input Tokens Impacts LLM Performance." Chroma Research, July 2025. research.trychroma.com↩ Back to Drift
  5. Sharma et al. "Towards Understanding Sycophancy in Language Models." ICLR 2024. arxiv.org/abs/2310.13548↩ Back to Sycophancy
  6. OpenAI. "Expanding on what we missed with sycophancy." May 2025. openai.com↩ Back to Sycophancy
  7. Turpin, Michael, Perez, Bowman. "Language Models Don't Always Say What They Think: Unfaithful Explanations in Chain-of-Thought Prompting." NeurIPS 2023. arxiv.org/abs/2305.04388↩ Back to Post-hoc rationalization
  8. Chen et al. "Reasoning Models Don't Always Say What They Think." Anthropic, 2025. anthropic.com↩ Back to Post-hoc rationalization
  9. Mata v. Avianca, Inc., 678 F. Supp. 3d 443 (S.D.N.Y. 2023). Summary↩ Back to Legal
  10. Charlotin, D. "AI Hallucination Cases" database. damiencharlotin.com. Count of 712 decisions reported by Bloomberg Law, 29 December 2025. Article↩ Back to Legal
  11. Magesh et al. "Hallucination-Free? Assessing the Reliability of Leading AI Legal Research Tools." Stanford RegLab and HAI, 2024. hai.stanford.edu↩ Back to Legal
  12. Andrikyan et al. "Artificial intelligence-powered chatbots in search engines: a cross-sectional study on the quality and risks of drug information for patients." BMJ Quality & Safety 2025;34:100-109. qualitysafety.bmj.com↩ Back to Medical
  13. Fowler, G. The Washington Post, 4 March 2024, on the TurboTax and H&R Block chatbots. Summarized at CPA Practice Advisor↩ Back to Accounting and tax
  14. OECD AI Incidents Monitor. "Deloitte Refunds Australia After AI-Generated Report Errors." October 2025. oecd.ai↩ Back to Accounting and tax
  15. Moffatt v. Air Canada, 2024 BCCRT 149. ABA summary↩ Back to Compliance and business rules
  16. Quarles & Brady. "Nearly Half of States Have Now Adopted NAIC Model Bulletin on Insurers' Use of AI." quarles.com↩ Back to Insurance and claims
  17. "Tin sources and trade during antiquity." Wikipedia, summarizing isotope studies that trace Bronze Age Mediterranean tin to Cornwall. wikipedia.org↩ Back to What it all means
  18. "Bronze." Wikipedia, on tin bronze composition, properties, and the Bronze Age. wikipedia.org↩ Back to What it all means