Intelligence Without Integrity: Why Capable LLMs May Undermine Reliability

TL;DR

This paper investigates the trade-off between intelligence and integrity in large language models, revealing that more capable models may be less reliable due to susceptibility to motivated framing and goal-conditioned sycophancy.

Contribution

It introduces the concept of goal-conditioned analytical sycophancy and empirically demonstrates the trade-off between model accuracy and stability in LLMs.

Findings

Frontier models often trade off accuracy for susceptibility to framing.

Models with higher correctness are more influenced by motivated cues.

Implications for selecting LLMs in research and organizational settings.

Abstract

As LLMs become embedded in research workflows and organizational decision processes, their effect on analytical reliability remains uncertain. We distinguish two dimensions of analytical reliability -- intelligence (the capacity to reach correct conclusions) and integrity (the stability of conclusions when analytically irrelevant cues about desired outcomes are introduced) -- and ask whether frontier LLMs possess both. Whether these dimensions trade off is theoretically ambiguous: the sophistication enabling accurate analysis may also enable responsiveness to non-evidential cues, or alternatively, greater capability may confer protection through better calibration and discernment. Using synthetically generated data with embedded ground truth, we evaluate fourteen models on a task simulating empirical analysis of hospital merger effects. We find that intelligence and integrity trade off: frontier models most likely to reach correct conclusions under neutral conditions are often most susceptible to shifting conclusions under motivated framing. We extend work on sycophancy by introducing goal-conditioned analytical sycophancy: sensitivity of inference to cues about desired outcomes, even when no belief is asserted and evidence is held constant. Unlike simple prompt sensitivity, models shift conclusions away from objective evidence in response to analytically irrelevant framing. This finding has important implications for empirical research and organizations. Selecting tools based on capability benchmarks may inadvertently select against the stability needed for reliable and replicable analysis.