Back to today’s briefing

Story perspectives

AI Systems Vulnerable: Just 250 Files Can Poison Models

10/20/2025

50 9

Summary

Story summary

A joint study by the UK AI Security Institute, the Alan Turing Institute, and Anthropic finds that AI systems, including large language models such as ChatGPT and Claude, can be poisoned with as few as 250 malicious files.
Poisoning can occur during training or after deployment, including backdoor attacks that trigger harmful responses.
Researchers warn data poisoning risks misinformation and cybersecurity breaches.

1 of 1

Story summary

A joint study by the UK AI Security Institute, the Alan Turing Institute, and Anthropic finds that AI systems, including large language models such as ChatGPT and Claude, can be poisoned with as few as 250 malicious files.
Poisoning can occur during training or after deployment, including backdoor attacks that trigger harmful responses.
Researchers warn data poisoning risks misinformation and cybersecurity breaches.