Automatically assuming other people are being honest, unless there's strong evidence of deception.
Imagine you have a friend who almost always tells you the truth. Because they're honest nearly every time, you stop even thinking about whether they might be lying—you just believe them automatically. That's how your brain works with everyone: it assumes people are telling the truth until something really weird happens that makes you go 'Wait a minute...' It's like your brain has a setting stuck on 'believe' until an alarm goes off.
Truth-Default Theory posits that humans passively accept incoming communication as honest unless a specific trigger activates suspicion and shifts them out of this default state. This is not mere gullibility but an adaptive cognitive stance: because the vast majority of everyday communication is truthful, defaulting to belief yields correct judgments far more often than systematic skepticism would. The theory further explains the 'veracity effect'—the finding that people are significantly better at correctly identifying truths than at correctly identifying lies—as a direct consequence of this truth-biased default. Critically, the truth-default can be disrupted by specific triggers such as discovering a motive to deceive, encountering behavioral anomalies, or receiving third-party warnings, at which point people shift into an active assessment mode where they weigh evidence for and against honesty.
The same glitch looks different depending on the terrain. Finance, medicine, a relationship, a team — same mechanism, different costume.
Investors tend to accept financial advisors' and fund managers' reported performance figures at face value, not because they've evaluated the evidence and deemed it credible, but because the possibility of deception simply doesn't arise. This makes them vulnerable to Ponzi schemes and fraudulent reporting.
Patients passively accept a physician's diagnosis and treatment recommendations without considering that the doctor might be wrong or influenced by pharmaceutical incentives. Clinicians similarly tend to accept patients' self-reported symptoms as truthful, potentially delaying the identification of drug-seeking behavior or factitious disorders.
Timothy R. Levine, first presented at the National Communication Association in 2012, formally published in the Journal of Language and Social Psychology in 2014, and fully elaborated in his 2019 book 'Duped.' The concept of truth-bias itself was first coined by Steven McCornack and Malcolm Parks in 1986.
Cooperative communication was essential for early human survival—sharing information about food sources, predators, and social alliances required that group members could quickly and efficiently process others' statements as true. A species-wide default toward belief enabled rapid information transfer and social coordination, providing a massive collective advantage. Since most communication within small, interdependent groups was indeed honest (deception carried high reputational costs), the truth-default yielded correct inferences the vast majority of the time.
Large language models and AI chatbots inherit a form of truth-default from their training data—they tend to treat input prompts as sincere and factual, making them vulnerable to prompt injection attacks where malicious instructions are disguised as legitimate queries. Similarly, AI content detectors struggle because the systems default to classifying text as human-written unless strong stylistic triggers indicate otherwise, mirroring the human truth-default. Users also apply truth-default to AI outputs, passively accepting generated text as accurate without verifying factual claims.
This page taught you the name. The deck turns the name into reflex. 1,100+ swipeable scenarios, 1,100+ defenses, 650+ detection prompts — spaced-repetition Swipe Deck, unlimited Spot-the-Bias Quiz, Defense Playbook, Pre-Flight, My Blindspots, Cheat Sheets, Field Guide e-book. $39.53$59.
