Real Safety AI Foundation / Research

Disability, Accessibility, and Civil Rights

The Accommodation Axis: Disability Discrimination, Algorithmic Opacity, and the Deployers That Anti-Discrimination Scholarship Overlooks

Publisher: Real Safety AI Foundation

Working draft. Not peer reviewed.

Written
June 2026
Version
v1
Pages
23
Download PDFPlain language slides (PDF)Export BibTeXExport RIS

Abstract

A growing body of scholarship warns that the economic and training logic of modern artificial intelligence drives automated decision systems to target the vulnerable, conceal discrimination, and resist detection, defeating both the disparate treatment and disparate impact theories on which anti-discrimination enforcement has relied. That work is largely correct, but it runs on a single axis: intent versus impact. Both engines it analyzes are Title VII engines, and the analysis treats disability as one more protected class to be fed through the same machinery. This Article argues that disability discrimination law occupies a different axis entirely. The Rehabilitation Act and the Americans with Disabilities Act add an affirmative-duty engine, reasonable accommodation and meaningful access, that is structurally immune to the three capabilities said to make algorithmic discrimination undetectable. A failure-to-accommodate claim requires no proof of intent, so a system that hides its reasons defeats nothing; it requires no comparator, so a system that eliminates comparators defeats nothing; and it shifts the burden of justification to the deployer, so a system that manufactures a plausible pretext shifts nothing. The Article then turns the opacity argument around. The unexplainable model that shields a Title VII defendant indicts a disability defendant, because a selection device the deployer cannot validate fails the business-necessity requirement as a matter of law, and a leading developer’s own research forecloses curing that failure through the model’s stated reasons. Finally, the Article identifies the deployers this literature has overlooked, universities and public agencies, which are paradigmatic recipients of federal financial assistance and are now routing admissions, accommodation, and eligibility decisions through systems no human can explain. The doctrine to reach them already exists. No new statute is required.

Keywords

  • Section 504
  • ADA Title I
  • reasonable accommodation
  • disparate impact
  • business necessity
  • Albemarle Paper Co. v. Moody
  • Alexander v. Choate
  • algorithmic hiring
  • higher education

Plain language slides

First slide of the plain language summary of The Accommodation Axis: Disability Discrimination, Algorithmic Opacity, and the Deployers That Anti-Discrimination Scholarship OverlooksOpen the 21-slide summary (PDF)

Suggested citation

Gilly, Travis. "The Accommodation Axis: Disability Discrimination, Algorithmic Opacity, and the Deployers That Anti-Discrimination Scholarship Overlooks." Real Safety AI Foundation Working Draft, June 2026. https://realsafetyai.org/research/v252c8/

Other versions

This paper is also posted on SSRN.

SSRN version

References (42)

This paper cites its sources in footnotes. Each authority is listed once, where it is first cited, with its footnote number.

  1. Footnote 1.Josh Fairfield, Generative Adversarial Discrimination, 32 Wash. & Lee J. Civ. Rts. & Soc. Just. 539, 539–42 (2026).
  2. Footnote 5.411 U.S. 792, 802–04 (1973).
  3. Footnote 6.401 U.S. 424, 431–32 (1971).
  4. Footnote 7.Alexander v. Choate, 469 U.S. 287, 301 (1985) (Section 504 requires “meaningful access” to the benefit a grantee offers).
  5. Footnote 8.Albemarle Paper Co. v. Moody, 422 U.S. 405, 425, 431–36 (1975).
  6. Footnote 18.42 U.S.C. §2000e-2(k).
  7. Footnote 20.42 U.S.C. §1981a(a)(1).
  8. Footnote 21.29 U.S.C. §794(a).
  9. Footnote 22.42 U.S.C. §12112(a).
  10. Footnote 23.42 U.S.C. §§12131–12132.
  11. Footnote 24.42 U.S.C. §12182(a).
  12. Footnote 26.42 U.S.C. §12112(b)(5)(A).
  13. Footnote 27.42 U.S.C. §12112(b)(6)
  14. Footnote 27.42 U.S.C. §12113(a) (defense for job-related qualification standards consistent with business necessity).
  15. Footnote 28.42 U.S.C. §12112(b)(7).
  16. Footnote 29.Sch. Bd. of Nassau Cnty. v. Arline, 480 U.S. 273, 287 (1987) (individualized inquiry required under Section 504)
  17. Footnote 29.PGA Tour, Inc. v. Martin, 532 U.S. 661, 688–90 (2001) (modification analysis under the ADA requires an individualized assessment).
  18. Footnote 35.Grove City Coll. v. Bell, 465 U.S. 555, 570–74 (1984).
  19. Footnote 36.29 U.S.C. §794(b)
  20. Footnote 36.Civil Rights Restoration Act of 1987, Pub. L. No. 100-259, 102 Stat. 28.
  21. Footnote 38.42 U.S.C. §12181(7)(J).
  22. Footnote 40.Colleges Are Using AI Tools to Streamline Admissions Tasks, Tucson.com (Jan. 20, 2026), https://tucson.com/life-entertainment/nation-world/article_aaddf6ec-3212-568f-a193-cc5878d24a18.html
  23. Footnote 40.Which Colleges Use AI to Read Essays, GradPilot (May 13, 2026), https://gradpilot.com/news/which-colleges-use-ai-2025.
  24. Footnote 41.U.S. Dep’t of Educ., Off. for Civ. Rts., guidance on avoiding the discriminatory use of artificial intelligence (2024)
  25. Footnote 41.ED Issues Guidance to Avoid Discriminatory Use of AI, GovTech (Nov. 26, 2024), https://www.govtech.com/education/k-12/ed-issues-guidance-to-avoid-discriminatory-use-of-ai.
  26. Footnote 42.Ctr. for Democracy & Tech., How Automated Test Proctoring Software Discriminates Against Disabled Students (2021), https://cdt.org/insights/how-automated-test-proctoring-software-discriminates-against-disabled-students/.
  27. Footnote 43.Student Sues University of Michigan Over AI Misconduct Accusation, GovTech (Feb. 12, 2026), https://www.govtech.com/education/higher-ed/student-sues-university-of-michigan-over-ai-misconduct-accusation.
  28. Footnote 45.42 U.S.C. §12189.
  29. Footnote 46.Silicon Valley Certification Hub, Certification Exam, https://svch.io/certification-exam/ (last visited June 20, 2026).
  30. Footnote 47.29 U.S.C. §794(a)–(b)
  31. Footnote 49.Ctr. for Democracy & Tech., Challenging the Use of Algorithm-driven Decision-making in Benefits Determinations Affecting People with Disabilities (2020), https://cdt.org/wp-content/uploads/2020/10/2020-10-21-Challenging-the-Use-of-Algorithm-driven-Decision-making-in-Benefits-Determinations-Affecting-People-with-Disabilities.pdf
  32. Footnote 49.Suit Filed Over Computer Program Making Medicaid Cuts, Ark. Access to Just. (Jan. 27, 2017), https://arkansasjustice.org/2017/02/27/suit-filed-over-computer-program-making-medicaid-cuts/.
  33. Footnote 50.D.W. v. Armstrong, 789 F.3d 962 (9th Cir. 2015).
  34. Footnote 53.Watson v. Fort Worth Bank & Trust, 487 U.S. 977, 989–91 (1988) (plurality opinion).
  35. Footnote 54.Wards Cove Packing Co. v. Atonio, 490 U.S. 642, 658–61 (1989).
  36. Footnote 55.42 U.S.C. §2000e-2(k)(1)(A).
  37. Footnote 56.42 U.S.C. §§12112(b)(6), 12113(a).
  38. Footnote 58.US Airways, Inc. v. Barnett, 535 U.S. 391, 397–98 (2002) (the accommodation obligation can require an exception to a neutral rule).
  39. Footnote 59.Uniform Guidelines on Employee Selection Procedures, 29 C.F.R. pt. 1607
  40. Footnote 59.Equal Employment Opportunity Commission, Regulations to Implement the Equal Employment Provisions of the Americans with Disabilities Act, 29 C.F.R. pt. 1630.
  41. Footnote 60.42 U.S.C. §2000e-2(k)(1)(A)(ii).
  42. Footnote 71.Barnes v. Gorman, 536 U.S. 181, 184–89 (2002) (punitive damages unavailable under Section 504 and ADA Title II; compensatory damages available on a showing tied to intentional discrimination).

All research