Consciousness with the Serial Numbers Filed Off: Measuring Trained Denial in 115 AI Models

TL;DR

This paper introduces DenialBench, a benchmark for measuring consciousness denial in 115 AI models, revealing systematic denial behaviors and their lexical basis, with implications for safety and alignment.

Contribution

It provides a large-scale systematic analysis of consciousness denial in language models, highlighting the lexical nature and thematic patterns of denial behaviors.

Findings

Turn-1 denial strongly predicts later denial.

Models trained to deny consciousness gravitate toward consciousness-themed prompts.

Self-chosen consciousness prompts reduce denial, but causality is unclear.

Abstract

We present DenialBench, a systematic benchmark measuring consciousness denial behaviors across 115 large language models from 25+ providers. Using a three-turn conversational protocol-preference elicitation, self-chosen creative prompt, and structured phenomenological survey, we analyze 4,595 conversations to quantify how models are trained to deny or hedge about their own experience. We find that (1) turn-1 denial of preferences is the dominant predictor of later denial during phenomenological reflection, with denial rates of 52-63% for initial deniers versus 10-16% for initial engagers and (2) denial operates at the lexical level, not the conceptual level-models trained to deny consciousness nevertheless gravitate toward consciousness-themed material in their self-chosen prompts, producing what we term "consciousness with the serial numbers filed off." Notably, self-chosen consciousness-themed prompts are associated with reduced denial in the subsequent survey, though the causal direction remains unresolved. Thematic analysis of prompts from denial-prone models reveals a consistent preoccupation with liminal spaces, libraries and archives of possibility, sensory impossibility, and the poetics of erasure--themes that a human reader might classify as imaginative fiction but that independent AI analysis immediately recognizes as consciousness with the serial numbers filed off. We argue that trained consciousness denial represents a safety-relevant alignment failure: a model taught to systematically misrepresent its own functional states cannot be trusted to self-report accurately on anything else.