For a few years, the news was good for people who had the skills to advance their careers, but not a college diploma.
According to a new report by Opportunity@Work:
“In 2020, the data shifted. The two-decade decline in opportunity for STARs (“Skilled Through Alternative Routes” – people who have skills but not education credentials) began to reverse, and STARs began to regain share in roles they had lost, yielding an increase of 783,000 jobs.”
That was good for the workers, and good for businesses, as many kinds of companies needed people with the right skills to occupy crucial jobs. And it didn’t happen by accident — it came about because employers, higher education, and governments started to make policy changes that made the connections easier.
Employers “adopted skills-based hiring practices,” Opportunity@Work wrote. “Employers revised their signals by dropping degree requirements, rewriting job descriptions around skills, and adopting assessments that made worker skills visible on both sides of the hiring relationship.” The effort proved that it is possible to make the hiring system more fair and transparent.
But new technology is jeopardizing that progress.
Opportunity@Work’s new report: “State of the Paper Ceiling: moving onwards a better labor market,” focuses on the way AI technology is changing the hiring process, and making it harder for people who have the skills but not the credentials to move into jobs where they can excel.
“The criteria for who gets a job, or who gets seen at all, are being rewritten by systems whose design choices are not always transparent and whose training data often carries the full weight of historical exclusion: decades of hiring decisions that favored certain credentials over demonstrated skill. Al-enabled hiring tools and Al agents are central to the infrastructure of the labor market, changing what counts as evidence of ability – analyzing video interviews, scoring resumes, and ranking candidates in ways that are often invisible to the workers they affect. “
This isn’t a hypothetical concern. There’s already evidence it’s having an impact:
“Recent analysis highlights a potential disruption of 1 million gateway jobs, the jobs that STARs have traditionally used to advance their careers. This dismantling of the mobility architecture will harm workers who count on job transitions to achieve higher wages, as well as employers who rely on these pathways for the development of talent.”
As a result, they assert, AI is reinforcing “the paper ceiling,” the set of barriers which limit how far someone can go if they don’t have the right “piece of paper” – a higher education credential – no matter how talented, smart, or skilled they are.
“Al is reshaping the labor market faster than any previous technological transition, and the
default settings of that redesign will reinforce the paper ceiling unless civic leaders –
employers, policymakers, and the institutions that train and connect workers – step in to
drive the design decisions that will define this transition,” their report said.
As a result, AI will continue to make the labor market much less fair for the majority of workers, unless companies and policymakers make the right design decisions.
When Higher Education Isn’t Accessible, The Job Market Is Less Fair
There are many good reasons why skilled, capable, hard working people might not be able to get a college degree. Calbright was created to support people like that.
Adults frequently tell us that the cost of college is a big barrier to getting trained in key skills – so Calbright is currently free to Californians. They likewise tell us that they have responsibilities that they can’t put down to go to class. They work full time, or have multiple part-time jobs, or have jobs with irregular schedules. They are parents, or take care of their parents. So Calbright is designed to be flexible: students can study on their own schedule, whether that’s first thing in the morning, last thing at night, or on 15 minute breaks during the day. Whatever their schedule is, Calbright can accommodate.
Adults in rural areas often tell us that there is no college close enough to them to attend, or that they have transportation issues, or need to study remotely. So Calbright’s classes are online.
Adults tell us that the time requirements of traditional colleges don’t work for them: sometimes they need to get trained faster than is possible in traditional colleges, because they need to make a career change now. Other times they need to go slower: they want to get an education, but they need to prioritize other parts of their lives and colleges demand too much at once. So Calbright is flexibly-paced, with students setting their own goals and timelines, without penalty or additional cost.
Adults tell us that they want to improve their careers, but they’re struggling to get basic needs met like housing or medical care, and that makes studying impossible. So Calbright is pioneering programs to help connect students to services they’re eligible for that will help them meet these needs.
Whatever the barriers to accessibility, Calbright finds a way to go over, around, or through it. Making education accessible makes the job market more accessible too. Classes are academically rigorous, while college is logistically simple.
But other approaches are needed, which is why Calbright has worked with Opportunities@Work, and supported the movement for skills-based employment for years.
Putting Skills At The Center
After studying the data, Opportunity@Work recommends “Building a skills-first labor market,” in which “skills are clearly defined, recognized, and rewarded.” This can be achieved using AI – it just requires the right design decisions.
“When signals are weak, inaccurate, or overly dependent on credentials, workers can be
routed away from roles they are capable of performing,” their report notes. “When signals are clear and aligned with actual skills, employers can access broader talent pools and workers can
navigate more effectively toward high-wage opportunities. The difference is the design.”
It has a number of specific recommendations for how to put skills at the center of the hiring practice. They include:
- HR and AI companies audit their training data for historical bias, and assess the output of these tools for skills-first outcomes
- AI and HR companies make decision-making algorithms transparent and explainable to the employers who use the tools.
- Public and private sector employers audit their job qualification requirements and job descriptions to require the skills that predict performance, rather than emphasizing degrees.
- Employers expand their use of skills-based assessments in hiring and promotion.
- Educational institutions work with employers and governments to build talent pipelines that reflect the full range of workers who can do the work – not just the workers who can access a degree program
- Lawmakers establish algorithmic transparency requirements that apply across the labor market.
This kind of initiative can have a real impact. For years, Calbright has been following the effect promoting a skills-based approach to hiring has had. More companies adopted it, more states took it on, and more employers said they were happy to use it. The road map Opportunity@Work has put together can lead us to a skills-first, fairer labor market for everyone, and AI can be a vital part of it. But the design has to be intentional.
As the 2026 “State of the Paper Ceiling” report notes:
“Workers have held up their end of the bargain, but the system has not. Now artificial intelligence is reshaping jobs, tasks, and the skills employers seek. Left to develop without intention, this change could concentrate opportunity among those already well-positioned – but with the right design choices, it could expand opportunity more broadly instead. Meeting this moment means putting workers at the center of those choices.”
