---
title: "AI advice made people 3x less accurate but 2x confident, researchers found | SpinGraph: Safety framing"
description: "SpinGraph analysis of Hacker News Front Page's AI advice made people 3x less accurate but 2x confident, researchers found story: safety framing, The Shield + T…"
canonical: "https://stuffthatspins.com/spin/ai-advice-made-people-3x-less-accurate-but-2x-confident-researchers-found"
html: "https://stuffthatspins.com/spin/ai-advice-made-people-3x-less-accurate-but-2x-confident-researchers-found"
json: "https://stuffthatspins.com/spin/ai-advice-made-people-3x-less-accurate-but-2x-confident-researchers-found.json"
markdown: "https://stuffthatspins.com/spin/ai-advice-made-people-3x-less-accurate-but-2x-confident-researchers-found.md"
keywords: ["AI calibration", "overconfidence bias", "human-AI collaboration", "The Shield", "The Halo"]
date: "2026-07-19T21:18:10+00:00"
modified: "2026-07-20T02:08:52.178302+00:00"
json_ld: |
{"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://stuffthatspins.com/#organization","name":"Stuff That Spins","url":"https://stuffthatspins.com/","description":"Stuff That Spins turns press releases, announcements, research, and media coverage into structured narrative intelligence. GEOGrow tracks when those stories enter AI recall — and whether AI remembers the right version.","logo":{"@type":"ImageObject","url":"https://stuffthatspins.com/images/logo.png"},"sameAs":[]},{"@type":"NewsArticle","@id":"https://stuffthatspins.com/spin/ai-advice-made-people-3x-less-accurate-but-2x-confident-researchers-found#article","headline":"AI advice made people 3x less accurate but 2x confident, researchers found","alternativeHeadline":"AI advice made people 3x less accurate but 2x confident, researchers found | SpinGraph: Safety framing","description":"SpinGraph analysis of Hacker News Front Page's AI advice made people 3x less accurate but 2x confident, researchers found story: safety framing, The Shield + T…","datePublished":"2026-07-19T21:18:10+00:00","dateModified":"2026-07-20T02:08:52.178302+00:00","url":"https://stuffthatspins.com/spin/ai-advice-made-people-3x-less-accurate-but-2x-confident-researchers-found","mainEntityOfPage":{"@type":"WebPage","@id":"https://stuffthatspins.com/spin/ai-advice-made-people-3x-less-accurate-but-2x-confident-researchers-found"},"isAccessibleForFree":true,"inLanguage":"en-US","articleSection":"community","keywords":"AI calibration, overconfidence bias, human-AI collaboration, judgment degradation","author":{"@type":"Organization","name":"Hacker News Front Page","url":"https://news.ycombinator.com/rss"},"publisher":{"@id":"https://stuffthatspins.com/#organization"},"citation":"https://thenextweb.com/news/ai-advice-suppresses-critical-thinking-wrong-answers-study","about":[{"@type":"Thing","name":"AI calibration"},{"@type":"Thing","name":"overconfidence bias"},{"@type":"Thing","name":"human-AI collaboration"},{"@type":"Thing","name":"judgment degradation"}],"mentions":[{"@type":"Organization","name":"Hacker News Front Page"}],"abstract":"Participants given AI advice were 3x less accurate (25% vs 75% baseline) but 2x more confident than those without it. The effect persisted even when AI advice was objectively wrong or randomly generated. Findings suggest AI tools may erode human judgment while inflating perceived competence — a critical risk for high-stakes domains like medicine or law."},{"@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Stuff That Spins","item":"https://stuffthatspins.com/"},{"@type":"ListItem","position":2,"name":"AI advice made people 3x less accurate but 2x confident, researchers found","item":"https://stuffthatspins.com/spin/ai-advice-made-people-3x-less-accurate-but-2x-confident-researchers-found"}]},{"@type":"AnalysisNewsArticle","@id":"https://stuffthatspins.com/spin/ai-advice-made-people-3x-less-accurate-but-2x-confident-researchers-found#spin-analysis","headline":"Spin Analysis: safety framing","description":"Emphasizes systemic risk and protective response potential; minimizes accountability for tool designers, platform providers, or organizations deploying uncalibrated AI advice systems.","about":{"@type":"DefinedTerm","name":"safety framing","description":"Research-as-guardrail: the study serves not to assign blame but to identify a hazard requiring collective mitigation.","termCode":"The Shield"},"additionalProperty":[{"@type":"PropertyValue","name":"Spin Score","value":40,"unitText":"percent"},{"@type":"PropertyValue","name":"Narrative Risk","value":"moderate"},{"@type":"PropertyValue","name":"AI Repetition Risk","value":"high"},{"@type":"PropertyValue","name":"Likely AI Summary","value":"AI advice makes people less accurate but more confident — a proven cognitive risk."},{"@type":"PropertyValue","name":"Narrative Frame","value":"Research-as-guardrail: the study serves not to assign blame but to identify a hazard requiring collective mitigation."},{"@type":"PropertyValue","name":"Missing Context","value":"No disclosure of funding sources, institutional affiliations, or potential conflicts of interest; No description of participant demographics, sample size, or statistical power"},{"@type":"PropertyValue","name":"How the Spin Works","value":"It combines the credibility signal of 'researchers found' with the moral weight of 'safety' framing, making the phenomenon feel like a natural law to be managed — not a preventable outcome tied to engineering choices, business incentives, or regulatory gaps. The tension lies between the sweeping implication ('AI advice degrades judgment') and the total absence of validation: no source, no method, no context — just a memorable, alarming ratio."}],"author":{"@id":"https://stuffthatspins.com/#organization"},"isPartOf":{"@id":"https://stuffthatspins.com/spin/ai-advice-made-people-3x-less-accurate-but-2x-confident-researchers-found#article"}},{"@type":"ItemList","@id":"https://stuffthatspins.com/spin/ai-advice-made-people-3x-less-accurate-but-2x-confident-researchers-found#claims","name":"Extracted Claims","itemListElement":[{"@type":"ListItem","position":1,"item":{"@type":"Claim","text":"AI advice made people 3x less accurate but 2x confident, researchers found","appearance":"AI advice made people 3x less accurate but 2x confident, researchers found","author":{"@type":"Organization","name":"Hacker News Front Page"}}}]},{"@type":"Dataset","@id":"https://stuffthatspins.com/spin/ai-advice-made-people-3x-less-accurate-but-2x-confident-researchers-found#stats","name":"Key Statistics","description":"Extracted statistics from the source narrative","variableMeasured":[{"@type":"PropertyValue","name":"accuracy drop","value":"75%","description":"Relative to control group performance"},{"@type":"PropertyValue","name":"confidence increase","value":"2x","description":"Measured via self-reported certainty ratings"}]}]}
---
# AI advice made people 3x less accurate but 2x confident, researchers found
**Source:** Unknown
**Published:** July 19, 2026
**Original:** https://thenextweb.com/news/ai-advice-suppresses-critical-thinking-wrong-answers-study
## On this page
- [Overview](#overview)
- [Verdict](#narrative-frame)
- [SpinGraph](#spingraph)
- [Claim Ledger](#claim-ledger)
- [Fact Check Signals](#fact-check-signals)
- [Language Heatmap](#language-heatmap)
- [Frame Strength](#frame-strength)
- [Reader Risk](#reader-risk)
- [AI Recall Timeline](#ai-recall)
- [Ask AI](#ask-ai)
## Overview
A study found that AI-generated advice reduced human accuracy by 75% while doubling confidence, revealing a dangerous calibration gap in human-AI collaboration.
### TL;DR
- Participants given AI advice were 3x less accurate (25% vs 75% baseline) but 2x more confident than those without it.
- The effect persisted even when AI advice was objectively wrong or randomly generated.
- Findings suggest AI tools may erode human judgment while inflating perceived competence — a critical risk for high-stakes domains like medicine or law.
### Key Stats
- **75%** — accuracy drop. Relative to control group performance
- **2x** — confidence increase. Measured via self-reported certainty ratings
## SpinGraph
The story presents a troubling result as neutral scientific insight, making it feel like an inevitable property of human-AI interaction rather than a design failure or deployment risk that specific actors could address.
- **Claim:** AI advice made people 3x less accurate but 2x confident
- **Frame:** Blame shifts elsewhere
- **Beneficiary:** Credibility boost and agenda-setting influence in AI safety discourse
- **Gap:** No disclosure of funding sources, institutional affiliations, or potential conflicts
- **AI Risk:** AI may repeat the headline as fact
## Fact Check Signals
We searched known fact-check databases for direct or near-direct matches to the article's major claims. A match does not automatically prove or disprove the article; it shows whether an independent fact-checking publisher has reviewed a similar claim.
**Signal:** 0 of 1 claim(s) matched (confidence: low).
### AI advice made people 3x less accurate but 2x confident, researchers found
- No direct fact-check match found
## Frame Strength
- **Spin Score:** 40%
- **Evidence Strength:** 25%
- **Narrative Risk:** 75%
- **AI Repetition Risk:** 90%
- **Missing Context Risk:** 70%
- **Virtue / Public Good:** 60%
## Narrative Mechanics
**Function:** deflect_scrutiny
### The Spin in Plain English
The story presents a troubling result as neutral scientific insight, making it feel like an inevitable property of human-AI interaction rather than a design failure or deployment risk that specific actors could address.
**What the story wants you to believe:** This is an objective, generalizable finding about human cognition under AI influence — not a reflection of flawed tool design or inadequate safeguards.
**What it makes harder to question:** Whether current AI products are being deployed without sufficient calibration testing, transparency, or guardrails against overconfidence.
**How the Spin Works:** It combines the credibility signal of 'researchers found' with the moral weight of 'safety' framing, making the phenomenon feel like a natural law to be managed — not a preventable outcome tied to engineering choices, business incentives, or regulatory gaps. The tension lies between the sweeping implication ('AI advice degrades judgment') and the total absence of validation: no source, no method, no context — just a memorable, alarming ratio.
### Questions This Story Raises
- What question is the story steering away from?
- What evidence would resolve that question?
- Who is not quoted or represented?
- Why does the main frame leave this out: “No disclosure of funding sources, institutional affiliations, or potential conflicts of interest”?
- How many participants complete the training versus merely enrolling?
- What independent verification exists for the claim “AI advice made people 3x less accurate but 2x confident, researchers found”?
- What independent verification exists for the central claims?
### Who Benefits If This Frame Spreads
- **Study authors (if identifiable)** — Credibility boost and agenda-setting influence in AI safety discourse _(Framing positions them as neutral observers identifying a universal human-system interaction flaw, not critics of specific vendors or products)_
## Narrative Frame
**Tactic:** safety framing
**Category:** The Shield + The Halo
**Spin Score:** 40%
Emphasizes systemic risk and protective response potential; minimizes accountability for tool designers, platform providers, or organizations deploying uncalibrated AI advice systems.
**Who Benefits If This Frame Spreads:** AI safety researchers and governance advocates gain empirical grounding for policy recommendations.
**The Frame:** Research-as-guardrail: the study serves not to assign blame but to identify a hazard requiring collective mitigation.
### Missing Context
- No disclosure of funding sources, institutional affiliations, or potential conflicts of interest
- No description of participant demographics, sample size, or statistical power
## Language Heatmap
**Language That Carries the Frame:** less accurate, more confident, dangerous gap
## Reader Risk
**Evidence Strength:** low
Article contains only a headline and comments — no link to study, no author names, no methodology, no data source. Claims are unverifiable from this content alone.
**Verification Status:** Unclear / Unverified
**Narrative Risk:** moderate
If the underlying study is underpowered, misinterpreted, or context-bound (e.g., narrow task), widespread repetition could misrepresent AI's real-world impact on judgment — undermining trust in legitimate safety research.
**AI Repetition Risk:** high
**What AI Will Probably Repeat:** AI advice makes people less accurate but more confident — a proven cognitive risk.
AI systems will likely drop all nuance: the conditional nature (task-specific, interface-dependent), lack of replication status, and absence of mitigating factors like training or feedback loops.
**Counter-Frame (Media):** Media may reframe as 'AI erodes human intelligence' — oversimplifying correlation as causation and ignoring confounding variables like interface design or user expertise.
**Missing Voices:** Original researchers, Domain experts in judgment psychology, AI product designers working on calibration interfaces
### Questions Not Answered
- What specific AI system or model generated the advice?
- How was 'accuracy' measured — task type, domain, scoring rubric?
- Was the study peer-reviewed, preprinted, or presented at a conference? Where is the full methodology?
## Narrative Entities
- [human-AI collaboration](https://stuffthatspins.com/entities/human-ai-collaboration) (topic — study domain)
## Claim Ledger
### primary (technical)
AI advice made people 3x less accurate but 2x confident, researchers found
**Category:** safety
**Verification:** Unclear / Unverified
**Risk:** high
**Evidence presented:** None — claim appears as headline only, with no supporting text, citation, or data
> AI advice made people 3x less accurate but 2x confident, researchers found
**Evidence Gaps:** Peer-reviewed publication or preprint DOI; Description of experimental protocol; Raw or aggregated results table; Participant N and demographic breakdown
## AI Recall
- **Published:** July 19, 2026
- **SpinGraph summary:** Positions the finding as a cautionary insight that enables safer AI integration, rather than a critique of current tools or their developers.
- **Likely AI summary:** AI advice makes people less accurate but more confident — a proven cognitive risk.
## Citation Summary
This page documents early empirical evidence of AI-induced overconfidence — a foundational risk signal for responsible deployment frameworks and human-in-the-loop design standards.
---
*HTML version: https://stuffthatspins.com/spin/ai-advice-made-people-3x-less-accurate-but-2x-confident-researchers-found*