LLMs may be more vulnerable to data poisoning than we thought

9 Oct 2025 · Alan Turing Institute – Blog UK

New evidence that LLMs are more vulnerable to data poisoning than assumed directly updates the risk baseline agencies should apply when procuring or deploying LLM-based tools.

Key points

Summary

A collaboration between the Alan Turing Institute, the UK AI Security Institute, and Anthropic is investigating data poisoning vulnerabilities in large language models. Early findings suggest LLMs may be more susceptible to this class of attack than previously understood. Data poisoning - where malicious inputs corrupt training or fine-tuning data to cause harmful or manipulated outputs - is a material supply chain risk for any agency deploying or procuring LLM-based services. The extracted text is brief and full findings have not yet been published.

Implications for Australian agencies

Implications are AI-generated. Starting points, not advice.