# Alignment Pretraining: AI Discourse Causes Self-Fulfilling (Mis)alignment

arXiv: https://arxiv.org/abs/2601.10160 · HTML: https://mlsanity.com/paper/2601.10160/

## Key facts

- Cited by 33 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 15 Jan 2026.

## Built on by (most cited first, 30 of 33)

- [Model Spec Midtraining: Improving How Alignment Training Generalizes](https://mlsanity.com/paper/2605.02087/) (2026)
- [Shaping capabilities with token-level data filtering](https://mlsanity.com/paper/2601.21571/) (2026)
- [Emergent and Subliminal Misalignment Through the Lens of Data-Mediated Transfer](https://mlsanity.com/paper/2605.12798/) (2026)
- [Gemma Needs Help: Investigating and Mitigating Emotional Instability in LLMs](https://mlsanity.com/paper/2603.10011/) (2026)
- [Evaluation Awareness in Language Models Has Limited Effect on Behaviour](https://mlsanity.com/paper/2605.05835/) (2026)
- [Training Agents to Self-Report Misbehavior](https://mlsanity.com/paper/2602.22303/) (2026)
- [Negation Neglect: When models fail to learn negations in training](https://mlsanity.com/paper/2605.13829/) (2026)
- [Positive Alignment: Artificial Intelligence for Human Flourishing](https://mlsanity.com/paper/2605.10310/) (2026)
- [Alignment midtraining for animals](https://mlsanity.com/paper/2604.13076/) (2026)
- [Less is More: Early Stopping Rollout for On-Policy Distillation](https://mlsanity.com/paper/2605.27028/) (2026)
- [When Role-playing, Do Models Believe What They Say?](https://mlsanity.com/paper/2606.11502/) (2026)
- [Persona Cartography: Charting Language Model Personality Traits in Weight Space](https://mlsanity.com/paper/2607.07916/) (2026)
- [Constitutional Midtraining: Content Presence Drives Alignment Gains](https://mlsanity.com/paper/2607.26654/) (2026)
- [Training on Documents About Monitoring Leads to CoT Obfuscation](https://mlsanity.com/paper/2605.15257/) (2026)
- [Reinforcement Learning Can Amplify Emergent Misalignment from Harmless Rewards](https://mlsanity.com/paper/2605.31328/) (2026)
- [Emergent Misalignment Recruits a Pre-existing Persona Subspace](https://mlsanity.com/paper/2607.21356/) (2026)
- [Your AI Travel Agent Would Book You a Bullfight: An Agentic Benchmark for Implicit Animal Welfare in Frontier AI Models](https://mlsanity.com/paper/2606.18142/) (2026)
- [Beyond Safe Data: Pretraining-Stage Alignment with Regular Safety Reflection](https://mlsanity.com/paper/2606.19168/) (2026)
- [Epistemic Goggles: A Pretrained Module that Induces an Epistemic Frame via Gradient Editing](https://mlsanity.com/paper/2607.01690/) (2026)
- [Data Attribution of Emergent Misalignment with Persona Features](https://mlsanity.com/paper/2608.11025/) (2026)
- [Synthetic Persona Pretraining: Alignment from Token Zero](https://mlsanity.com/paper/2608.13482/) (2026)
- [Self-Generated Text Recognition: Quality Heuristics, Cross-Task Transfer, and Downstream Bias in LLM Evaluation](https://mlsanity.com/paper/2608.26159/) (2026)
- [Sustaining AI safety: Control-theoretic external impossibility, intrinsic necessity, and structural requirements](https://mlsanity.com/paper/2605.12963/) (2026)
- [Understanding Goal Generalisation in Sequential Reinforcement Learning](https://mlsanity.com/paper/2605.23565/) (2026)
- [What Shapes Emergent Misalignment? Insights from Training Dynamics, Model Priors, and Data](https://mlsanity.com/paper/2606.20814/) (2026)
- [Helpfulness Hurts: Domain-Dependent Degradation of Mid-Trained Compassion Values Under Post-Training](https://mlsanity.com/paper/2606.26102/) (2026)
- [Demonstrating Generalization Failures via Mixtures of Conditional Policies](https://mlsanity.com/paper/2607.03478/) (2026)
- [The Assistant's Ideal Self](https://mlsanity.com/paper/2609.00304/) (2026)
- [Story Imprinting: AI Assistants Absorb Traits from Human Characters They Resemble](https://mlsanity.com/paper/2609.10883/) (2026)
- [Can Data Attribution Filter Out Subliminal Learning? Not Reliably](https://mlsanity.com/paper/2609.20027/) (2026)

## Abstract (arXiv)

Pretraining corpora contain extensive discourse about AI systems, yet the causal influence of this discourse on downstream alignment remains poorly understood. If prevailing descriptions of AI behaviour are predominantly negative, LLMs may internalise corresponding behavioural priors, giving rise to self-fulfilling misalignment. This paper provides the first controlled study of this hypothesis by pretraining 6.9B-parameter LLMs with varying amounts of (mis)alignment discourse. We find that discussion of AI contributes to misalignment. Upsampling synthetic training documents about AI misalignment leads to a notable increase in misaligned behaviour. Conversely, upsampling documents about aligned behaviour reduces misalignment scores from 45% to 9%. We consider this evidence of self-fulfilling alignment. These effects are dampened, but persist through post-training. Our findings establish the study of how pretraining data shapes alignment priors, or alignment pretraining, as a complement to post-training. We recommend practitioners consider pretraining for alignment alongside capabilities. We share our models, data, and evaluations at AlignmentPretraining.ai.