🛰️ Daily AI Frontier
‹ back to 2026-07-16

Pretraining Data Can Be Poisoned through Computational Propaganda

Research LLMs & Foundation Models

Merged summary

TL;DR - This paper shows that LLM pretraining corpora can be poisoned at web scale by injecting adversarial content through public discussion interfaces, and introduces a method to estimate whether such content survives web crawling and data curation.

  • Argues prior poisoning work relied on narrow sources (e.g., Wikipedia) that don't reflect the scale/heterogeneity of real pretraining data and ignored curation pipelines.
  • Demonstrates public discussion interfaces as a viable web-scale content injection vector for poisoning attacks.
  • Introduces HalfLife, an analysis for estimating adversarial content inclusion in web-crawl-based training data.
  • Establishes third-party webpage content as a plausible attack surface for LM pretraining, stressing the need to verify whether poison injections actually reach training data.

Sources (1)

Pretraining Data Can Be Poisoned through Computational Propaganda

arXiv cs.AI Victoria Graf, Hannaneh Hajishirzi, Noah A. Smith, David Kohlbrenner, Kyle Lo 2026-07-16 arXiv:2607.15267

TL;DR - This paper shows that LLM pretraining corpora can be poisoned at web scale by injecting adversarial content through public discussion interfaces, and introduces a method to estimate whether such content survives web crawling and data curation.

  • Argues prior poisoning work relied on narrow sources (e.g., Wikipedia) that don't reflect the scale/heterogeneity of real pretraining data and ignored curation pipelines.
  • Demonstrates public discussion interfaces as a viable web-scale content injection vector for poisoning attacks.
  • Introduces HalfLife, an analysis for estimating adversarial content inclusion in web-crawl-based training data.
  • Establishes third-party webpage content as a plausible attack surface for LM pretraining, stressing the need to verify whether poison injections actually reach training data.
item →