The future of DEI should be judged by whether people can get hired, do their work and progress without avoidable barriers. For disabled people, that means accessible systems, effective accommodations and a say in decisions. A new name for the department will achieve little if those conditions stay the same.
In September 2026, the US Equal Employment Opportunity Commission announced a lawsuit against the operators of a Las Vegas call centre. Among its allegations: a worker who needed short breaks for muscle spasms was told by a manager that he “didn’t do accommodations.” Another employee, who worked remotely and did not have use of his legs, was allegedly required to travel to the office for administrative tasks.
Those are allegations in an EEOC lawsuit, not findings by a court. They nevertheless identify the kind of question any serious discussion of inclusion has to confront. Who can change a working arrangement? What happens when someone asks? Who bears the cost of a refusal?
The public argument about diversity, equity and inclusion can feel several floors above that conversation. Executives debate terminology. Politicians attack an acronym. Consultants offer its replacement. Meanwhile, someone still needs an answer about a break.
The backlash has changed the conditions for DEI
When this article first appeared in March 2025, much of the argument concerned which companies would defend their diversity commitments and which would retreat. That remains relevant, but a list of corporate statements cannot tell us whether a workplace is fair.
The US federal campaign against DEI has also developed beyond the first announcements of Donald Trump’s second term. His March 26, 2026 executive order on federal contractors directed agencies to introduce contract terms prohibiting what the administration defines as racially discriminatory DEI activities. It contemplated consequences including contract termination and debarment.
The administration’s characterisation of DEI is a political and legal position, not a neutral description of everything done under that name. The order has been challenged: in June, a coalition of state attorneys general sued over its implementation. Those developments establish pressure and uncertainty; they do not support a blanket declaration that inclusion work is now illegal.
The EEOC’s own guidance on DEI-related discrimination says DEI is not defined in Title VII. The substance of an employment practice matters. Changing its label does not settle whether it is lawful.
Nor does the backlash make disability access optional. Covered US employers’ accommodation duties under the ADA continue to matter. Other countries have their own laws and political conditions. An American executive order is not a global instruction to abandon disabled workers.
Defending inclusion does not require defending every programme
One of the more useful contributions to the current debate comes from Lily Zheng. In a January 2026 excerpt from Fixing Fairness, Zheng criticises organisations that rely on isolated training sessions while leaving the working environment intact. Their alternative emphasises results, organisational change, broader participation and shared gains.
That criticism deserves attention from people who support DEI. A budget spent is not a barrier removed. Attendance at a workshop tells us that people attended a workshop. It does not tell us whether a disabled applicant could complete the assessment, or whether an employee’s accommodation survived a change of manager.
Still, “training does not work” is too blunt a conclusion. A 2025 field experiment by Cansin Arslan and colleagues, published in Science, found that timely, behaviourally designed training at a global telecommunications and engineering firm increased the shortlisting of women and non-national candidates, and the hiring of non-national candidates. That was a specific intervention in a specific setting, not a test of disability inclusion.
The distinction matters. An intervention should be assessed by what it asks people to do, when it reaches them and what changes afterwards. Neither buying training nor cancelling it is, by itself, an inclusion strategy.
I would rather see an organisation explain one barrier it has removed, with evidence, than announce another sweeping commitment it cannot account for.
Fairness has to reach the decisions that shape a working day
Iris Bohnet and Siri Chilazi’s Make Work Fair, published in 2025, puts organisational design at the centre of the discussion. Their argument directs attention towards the procedures through which employers recruit, evaluate and reward people, rather than expecting good intentions to overcome flawed systems.
That is a useful starting point for disability inclusion. Consider a hypothetical recruitment process that advertises flexibility but automatically rejects applicants who cannot drive. If driving is not an essential part of the role, the welcoming language has not fixed the screening rule.
Or consider a promotion process that rewards constant availability. An employee may deliver excellent work within agreed hours yet lose out to someone whose visibility is easier to notice. Before declaring the outcome meritocratic, the employer needs to explain what it is measuring.
These are decisions about the work itself. Our guide to writing inclusive job descriptions starts with that practical question: what does the person actually need to accomplish, and which requirements merely repeat an assumption?
A fairer process also needs someone with the authority to maintain it. An accessible application form will not compensate for inaccessible assessment software introduced six months later. Inclusion has to survive procurement, budget meetings and staff turnover.
A wider coalition must still recognise unequal barriers
Kenji Yoshino and David Glasgow offer a related, but distinct, response to the backlash. Discussing their 2026 book How Equality Wins in an NYU Law interview, they argue for a broader constituency for equality, room for disagreement and approaches that expand opportunity. Their discussion includes people whose concerns can be overlooked in conventional DEI programmes, including disabled people and older workers.
There is value in making the case for fair treatment intelligible beyond the people already persuaded. A colleague does not have to know the latest terminology to recognise that a selection process should test the skills a job requires.
But broader participation brings a test of its own. “Everyone belongs” can become an excuse to stop naming who encounters which barrier. Disability cannot be welcomed into the coalition and then made unspeakable when an accommodation requires a different arrangement.
An accessible default can benefit many people. Captions can help someone watching in a noisy room as well as a deaf colleague. That shared benefit is useful. It should not become a condition: a deaf employee’s access does not depend on proving that everybody else gains too.
The same applies when an accommodation costs money or helps only one person. If the argument for inclusion works only when there are no competing demands, it will fail precisely when someone needs it most.
US EEOC accommodation guidance illustrates the distinction between identical treatment and effective access. Covered employers generally must provide reasonable accommodations for qualified applicants and employees unless doing so would cause undue hardship. They need not remove essential job functions or lower uniformly applied production standards; an accommodation may help someone meet those standards.
Those are US rules, not a universal legal checklist. The editorial point travels further: an equal expectation does not always require an identical method. That is why inclusion without equity can leave the most consequential barriers untouched.
Disability cannot remain somebody else’s department
A June 2026 study by Lauren Shallish asks a revealing question: is disability a DEI issue? Drawing on eight interviews with US higher education administrators, faculty and staff, it examines how reporting structures, perceptions of cost and the handling of confidentiality can push disability to the margins of equity work.
This is a small qualitative study of higher education, not a representative survey of employers. Its value here is the organisational problem it makes visible: when disability is treated solely as an individual case for a specialist office, the wider institution can avoid examining its own arrangements.
A workplace should have a competent, confidential accommodation process. It should also ask why inaccessible tools were purchased, why meetings exclude some staff and why disabled employees have so little influence over the decisions that create those problems.
Those responsibilities can coexist. Protecting someone’s medical information does not prevent an organisation from involving disabled people in the design of its services. Participation should be voluntary, properly resourced and possible without turning a person’s diagnosis into public property.
Disabled people also have racial identities, genders, sexualities and class backgrounds. Giving disability a place in inclusion work should deepen those conversations. It should not create another competition over whose exclusion counts.
The United States is not the whole future
It is easy for an English-language debate dominated by American politics to imply that every organisation is moving in the same direction.
In August 2026, Australia’s ABC published its Diversity, Inclusion & Belonging Plan 2026–2029. The plan explicitly includes disability and sets out work on organisational practices, employee wellbeing and accessible content and services.
A published plan is evidence of an institutional commitment, not proof that the commitment has been achieved. It is nevertheless a useful reminder that retreat is not the only available response. The more demanding question is what an organisation will be able to demonstrate when it reports on the work.
What I would ask before believing a DEI success story
The future of diversity, equity and inclusion needs fewer declarations of victory and better questions. For disability inclusion, I would start with these:
- Can applicants use the hiring process? Examine the form, assessment and accommodation route, including what happens when the standard process does not work for someone.
- How long does effective support take? Record the time from a request to an arrangement that actually works, including delays and requests that disappear between teams.
- Who gets opportunities after being hired? Look at access to training, substantial assignments and promotion. Being employed is not the same as having a career.
- Can disabled staff influence decisions early enough? Consultation after a contract is signed or a policy is final leaves very little to influence.
- Who answers when the process fails? Give a named role the authority, time and resources to resolve barriers. A mailbox is not an accountable person.
These are editorial tests, not a validated scoring system. Read process data alongside employees’ accounts, protect privacy and avoid turning a small group into an identifiable dashboard. A difference in outcomes warrants investigation; it does not establish its own cause. Equally, tidy numbers cannot show everything people endured to produce them.
The commitment worth keeping
At URevolution, disability belongs at the centre of this argument. We should scrutinise programmes that fail disabled people and political efforts that make it easier to exclude them. Those positions are compatible.
The name an organisation chooses may change. Its responsibilities to the people who work there do not disappear with the old letterhead.
The future of DEI will be more credible when the person asking for an accessible interview, a workable adjustment or a fair opportunity can get a clear answer—and see it carried through.
