What If AI's Biggest Hiring Breakthrough Is Making More People Employable?

A delivery driver arrives at an apartment building with a customer's food. The address is incomplete. The customer isn't answering text messages. Normally, the driver would make a phone call, get directions, and complete the delivery.

But this driver is deaf and cannot communicate through a conventional voice call.

The driver knows how to do the job. Yet one communication requirement creates an obstacle that affects performance, earnings, and customer satisfaction.

Now imagine an AI tool that converts a typed message into a natural-sounding phone call.

The worker hasn't acquired a new qualification. The technology has simply removed an obstacle that was preventing them from demonstrating their existing abilities.

New research suggests that distinction can produce remarkable results.

When AI Removes the Barrier Instead of the Worker

In October 2026, the National Bureau of Economic Research published findings from a study examining how artificial intelligence affected deaf and hard-of-hearing delivery workers in China.

The researchers studied an AI communication system introduced in 2022 by a major food delivery platform. The technology allowed workers to send automated telephone messages to customers using text prompts.

Analyzing records for 4,284 deaf and hard-of-hearing workers and nearly 40,000 comparison workers, researchers found significant improvements following the technology's introduction.

Compared with changes among workers without hearing disabilities:

• Delivery times declined 3.5%.
• Late deliveries fell 11.2%.
• Negative customer ratings dropped 21.8%.
• Weekly hours worked increased 10.6%.
• Worker attrition declined 21%.

Completed orders also increased, and the earnings gap between deaf and hard-of-hearing workers and their colleagues narrowed.

The study examined one platform in China, and its findings cannot automatically be generalized to every occupation or disability. Nevertheless, it demonstrates something important.

AI didn't replace these workers. It helped them perform their jobs more effectively.

America's Overlooked Workforce

The potential implications for American employers are substantial.

According to the U.S. Bureau of Labor Statistics, the 2025 employment rate among working-age Americans with disabilities was 38.1%, compared with 74.8% for those without disabilities.

That gap reflects complex factors, including health circumstances, discrimination, inaccessible workplaces, and differences in available employment opportunities. Technology alone cannot resolve them all.

But the findings raise a question worth asking.

How many qualified people are being overlooked because employers confuse barriers to performing certain tasks with an inability to perform the underlying job?

Consider a customer-service position. The essential requirement is helping customers resolve problems. Does that always require speaking on a telephone, or could accessible communication tools make the work possible for someone who is otherwise qualified?

Similar questions apply to administrative, technical, and certain healthcare and manufacturing positions.

Under the Americans with Disabilities Act, covered employers generally must provide reasonable accommodations to qualified applicants and employees with disabilities unless doing so creates an undue hardship.

AI may offer additional ways to remove workplace obstacles while maintaining essential performance standards.

And as the delivery-worker study illustrates, accessibility can also produce measurable business benefits.

Recruiting Needs to Recognize Ability, Not Just Limitations

The implications extend beyond workplace accommodations.

Recruiting teams should examine whether their job requirements, assessments, and screening processes measure essential capabilities or simply reinforce conventional ways of performing work.

An applicant might be perfectly capable of handling a position with appropriate technology but struggle with an unnecessarily restrictive assessment.

This is especially important as employers introduce AI-powered hiring tools.

The EEOC and Department of Justice have warned that automated assessments can unlawfully exclude qualified applicants with disabilities when accessibility and reasonable accommodations are overlooked.

The technology intended to make recruiting more efficient should not create new barriers for the very people employers are trying to hire.

For talent acquisition leaders, that means evaluating candidates against essential job functions, ensuring assessments are accessible, and recognizing when technology could enable someone to perform a role successfully.

It also means taking a broader view of where qualified talent exists.

Finding the Talent Employers May Be Missing

This is where better talent discovery becomes part of the equation.

ProvenBase helps employers search beyond conventional recruiting databases to identify professionals whose skills and experience may not be apparent through traditional sourcing.

Its Talent 360 technology helps recruiters understand how requirements involving location, education, experience, and skills may be narrowing their candidate pools.

That information can support a more thoughtful evaluation of hiring criteria and reveal qualified professionals who might otherwise be overlooked.

ProvenBase is not an accessibility or accommodation platform. Its contribution is helping employers discover talent beyond their usual search boundaries.

Combined with accessible hiring practices and workplace technology, broader talent discovery can help organizations make better use of the workforce already available to them.

What If We've Been Asking AI the Wrong Question?

Most conversations about artificial intelligence and employment focus on how many workers the technology might replace.

But the NBER findings suggest another question deserves attention.

How many more people could work successfully if technology removed barriers that have historically limited their opportunities?

The research doesn't prove that AI will solve employment disparities. It does show that a targeted application can improve performance, earnings, and workforce participation.

For employers facing difficult hiring challenges, that possibility deserves consideration.

The next major breakthrough in talent acquisition may not involve replacing people with technology. It may involve recognizing capabilities that conventional hiring practices have overlooked.

The future of work shouldn't be measured only by how much AI can do without people. It should also be measured by how much more people can accomplish with AI.

Looking for qualified talent beyond the usual recruiting channels? Explore ProvenBase and discover what your current search may be missing.