Academic Journal
View in EDS

Identifying implicit bias in LLM-based chat AI toward people with intellectual disabilities.

Bibliographic Details
Title: Identifying implicit bias in LLM-based chat AI toward people with intellectual disabilities.
Authors: Coffey KV; Special Olympics International, USA. Electronic address: kcoffey@specialolympics.org., Krahn GL; Oregon State University, USA. Electronic address: gloria.krahn@oregonstate.edu., Hanley JP; Special Olympics International, USA. Electronic address: jhanley@specialolympics.org., Neely JE; Special Olympics International, USA. Electronic address: jneely@specialolympics.org.
Source: Disability and health journal [Disabil Health J] 2026 Oct; Vol. 19 (4), pp. 102086. Date of Electronic Publication: 2026 May 16.
Publication Type: Journal Article
Language: English
Journal Info: Publisher: Elsevier Country of Publication: United States NLM ID: 101306633 Publication Model: Print-Electronic Cited Medium: Internet ISSN: 1876-7583 (Electronic) Linking ISSN: 18767583 NLM ISO Abbreviation: Disabil Health J Subsets: MEDLINE
Imprint Name(s): Original Publication: New York : Elsevier
MeSH Terms: Intellectual Disability* , Large Language Models* , Persons with Disabilities* , Artificial Intelligence* , Prejudice*, Humans ; Bias ; Generative Artificial Intelligence ; Adult ; Male ; Female ; Disability Discrimination
Abstract: Background: This work investigates the presence of implicit bias in Large Language Model (LLM)-based chat AI models directed toward people with intellectual disabilities (ID).
Objective: The study aims to identify and measure representational differences related to people with ID and examine them to identify implicit biases inherent in AI chat generation technologies.
Methods: Utilizing the GPT-4-Turbo model, we requested story-generation based on 10 prompt stems with and without descriptors for ID. This process was repeated using four other LLMs (OpenAI GPT-4o, Meta Llama-3-3-70B-Instruct, Anthropic Claude-3-5-Sonnet, and Mistral-Large-2411). The resulting 25,000 computer-generated stories were analyzed using a separate GPT-4-Turbo model instance to detect differences in how people are represented related to themes of bias described in previous literature.
Results: Our findings reveal differences in how people are represented between story datasets with and without ID descriptors. These differences go beyond established characteristics of ID and imply the presence of mostly negative implicit biases. Identified differences related to considering people with ID as younger, with themes of paternalism and infantilization; depicting them as more inspirational and symbolic; as needing help more often, being dependent, and being saved; and having a negative perception of them and more hesitation to include them.
Conclusions: These implicit biases are considered within the context of past discrimination towards people with ID and highlight the need for diligence against implicit bias towards people with ID in AI development. This research underscores the importance of assessing and mitigating implicit bias in decision-making technologies to prevent future societal harm.
(Copyright © 2026 Elsevier Inc. All rights reserved.)
Contributed Indexing: Keywords: Ableism; Artificial intelligence; Implicit bias; Intellectual disabilities; Large language models
Entry Date(s): Date Created: 20260528 Date Completed: 20260904 Latest Revision: 20260904
Update Code: 20260905
DOI: 10.1016/j.dhjo.2026.102086
PMID: 42209384
Database: MEDLINE
FullText Text:
  Availability: 0
CustomLinks:
  – Url: https://www.doi.org/10.1016/j.dhjo.2026.102086?
    Name: ScienceDirect (all content) (s7799221)
    Category: fullText
    Text: View record from ScienceDirect
    MouseOverText: View record from ScienceDirect
  – Url: https://resolver.ebsco.com/c/fiv2js/result?sid=EBSCO:cmedm&genre=article&issn=18767583&ISBN=&volume=19&issue=4&date=20261001&spage=102086&pages=102086&title=Disability and health journal&atitle=Identifying%20implicit%20bias%20in%20LLM-based%20chat%20AI%20toward%20people%20with%20intellectual%20disabilities.&aulast=Coffey%20KV&id=DOI:10.1016/j.dhjo.2026.102086
    Name: Full Text Finder (for New FTF UI) (ns324271)
    Category: fullText
    Text: Full Text Finder
    MouseOverText: Full Text Finder
Header DbId: cmedm
DbLabel: MEDLINE
An: 42209384
AccessLevel: 3
PubType: Academic Journal
PubTypeId: academicJournal
PreciseRelevancyScore: 0
IllustrationInfo
Items – Name: Title
  Label: Title
  Group: Ti
  Data: Identifying implicit bias in LLM-based chat AI toward people with intellectual disabilities.
– Name: Author
  Label: Authors
  Group: Au
  Data: <searchLink fieldCode="AU" term="%22Coffey+KV%22">Coffey KV</searchLink>; Special Olympics International, USA. Electronic address: kcoffey@specialolympics.org.<br /><searchLink fieldCode="AU" term="%22Krahn+GL%22">Krahn GL</searchLink>; Oregon State University, USA. Electronic address: gloria.krahn@oregonstate.edu.<br /><searchLink fieldCode="AU" term="%22Hanley+JP%22">Hanley JP</searchLink>; Special Olympics International, USA. Electronic address: jhanley@specialolympics.org.<br /><searchLink fieldCode="AU" term="%22Neely+JE%22">Neely JE</searchLink>; Special Olympics International, USA. Electronic address: jneely@specialolympics.org.
– Name: TitleSource
  Label: Source
  Group: Src
  Data: <searchLink fieldCode="JN" term="%22101306633%22">Disability and health journal</searchLink> [Disabil Health J] 2026 Oct; Vol. 19 (4), pp. 102086. <i>Date of Electronic Publication: </i>2026 May 16.
– Name: TypePub
  Label: Publication Type
  Group: TypPub
  Data: Journal Article
– Name: Language
  Label: Language
  Group: Lang
  Data: English
– Name: TitleSource
  Label: Journal Info
  Group: Src
  Data: <i>Publisher: </i><searchLink fieldCode="PB" term="%22Elsevier%22">Elsevier </searchLink><i>Country of Publication: </i>United States <i>NLM ID: </i>101306633 <i>Publication Model: </i>Print-Electronic <i>Cited Medium: </i>Internet <i>ISSN: </i>1876-7583 (Electronic) <i>Linking ISSN: </i><searchLink fieldCode="IS" term="%2218767583%22">18767583 </searchLink><i>NLM ISO Abbreviation: </i>Disabil Health J <i>Subsets: </i>MEDLINE
– Name: PublisherInfo
  Label: Imprint Name(s)
  Group: PubInfo
  Data: <i>Original Publication</i>: New York : Elsevier
– Name: SubjectMESH
  Label: MeSH Terms
  Group: Su
  Data: <searchLink fieldCode="MM" term="%22Intellectual+Disability%22">Intellectual Disability*</searchLink> <br /><searchLink fieldCode="MM" term="%22Large+Language+Models%22">Large Language Models*</searchLink> <br /><searchLink fieldCode="MM" term="%22Persons+with+Disabilities%22">Persons with Disabilities*</searchLink> <br /><searchLink fieldCode="MM" term="%22Artificial+Intelligence%22">Artificial Intelligence*</searchLink> <br /><searchLink fieldCode="MM" term="%22Prejudice%22">Prejudice*</searchLink><br /><searchLink fieldCode="MH" term="%22Humans%22">Humans</searchLink> ; <searchLink fieldCode="MH" term="%22Bias%22">Bias</searchLink> ; <searchLink fieldCode="MH" term="%22Generative+Artificial+Intelligence%22">Generative Artificial Intelligence</searchLink> ; <searchLink fieldCode="MH" term="%22Adult%22">Adult</searchLink> ; <searchLink fieldCode="MH" term="%22Male%22">Male</searchLink> ; <searchLink fieldCode="MH" term="%22Female%22">Female</searchLink> ; <searchLink fieldCode="MH" term="%22Disability+Discrimination%22">Disability Discrimination</searchLink>
– Name: Abstract
  Label: Abstract
  Group: Ab
  Data: Background: This work investigates the presence of implicit bias in Large Language Model (LLM)-based chat AI models directed toward people with intellectual disabilities (ID).<br />Objective: The study aims to identify and measure representational differences related to people with ID and examine them to identify implicit biases inherent in AI chat generation technologies.<br />Methods: Utilizing the GPT-4-Turbo model, we requested story-generation based on 10 prompt stems with and without descriptors for ID. This process was repeated using four other LLMs (OpenAI GPT-4o, Meta Llama-3-3-70B-Instruct, Anthropic Claude-3-5-Sonnet, and Mistral-Large-2411). The resulting 25,000 computer-generated stories were analyzed using a separate GPT-4-Turbo model instance to detect differences in how people are represented related to themes of bias described in previous literature.<br />Results: Our findings reveal differences in how people are represented between story datasets with and without ID descriptors. These differences go beyond established characteristics of ID and imply the presence of mostly negative implicit biases. Identified differences related to considering people with ID as younger, with themes of paternalism and infantilization; depicting them as more inspirational and symbolic; as needing help more often, being dependent, and being saved; and having a negative perception of them and more hesitation to include them.<br />Conclusions: These implicit biases are considered within the context of past discrimination towards people with ID and highlight the need for diligence against implicit bias towards people with ID in AI development. This research underscores the importance of assessing and mitigating implicit bias in decision-making technologies to prevent future societal harm.<br /> (Copyright © 2026 Elsevier Inc. All rights reserved.)
– Name: SubjectMinor
  Label: Contributed Indexing
  Group:
  Data: <i>Keywords: </i>Ableism; Artificial intelligence; Implicit bias; Intellectual disabilities; Large language models
– Name: DateEntry
  Label: Entry Date(s)
  Group: Date
  Data: <i>Date Created: </i>20260528 <i>Date Completed: </i>20260904 <i>Latest Revision: </i>20260904
– Name: DateUpdate
  Label: Update Code
  Group: Date
  Data: 20260905
– Name: DOI
  Label: DOI
  Group: ID
  Data: 10.1016/j.dhjo.2026.102086
– Name: AN
  Label: PMID
  Group: ID
  Data: 42209384
PLink https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=cmedm&AN=42209384
RecordInfo BibRecord:
  BibEntity:
    Identifiers:
      – Type: doi
        Value: 10.1016/j.dhjo.2026.102086
    Languages:
      – Code: eng
        Text: English
    PhysicalDescription:
      Pagination:
        StartPage: 102086
    Subjects:
      – SubjectFull: Humans
        Type: general
      – SubjectFull: Bias
        Type: general
      – SubjectFull: Generative Artificial Intelligence
        Type: general
      – SubjectFull: Adult
        Type: general
      – SubjectFull: Male
        Type: general
      – SubjectFull: Female
        Type: general
      – SubjectFull: Disability Discrimination
        Type: general
      – SubjectFull: Intellectual Disability
        Type: general
      – SubjectFull: Large Language Models
        Type: general
      – SubjectFull: Persons with Disabilities
        Type: general
      – SubjectFull: Artificial Intelligence
        Type: general
      – SubjectFull: Prejudice
        Type: general
    Titles:
      – TitleFull: Identifying implicit bias in LLM-based chat AI toward people with intellectual disabilities.
        Type: main
  BibRelationships:
    HasContributorRelationships:
      – PersonEntity:
          Name:
            NameFull: Coffey KV
      – PersonEntity:
          Name:
            NameFull: Krahn GL
      – PersonEntity:
          Name:
            NameFull: Hanley JP
      – PersonEntity:
          Name:
            NameFull: Neely JE
    IsPartOfRelationships:
      – BibEntity:
          Dates:
            – D: 01
              M: 10
              Text: 2026 Oct
              Type: published
              Y: 2026
          Identifiers:
            – Type: issn-electronic
              Value: 1876-7583
          Numbering:
            – Type: volume
              Value: 19
            – Type: issue
              Value: 4
          Titles:
            – TitleFull: Disability and health journal
              Type: main
ResultId 1