# Why Unemployment Insurance Alone Is Not Enough in the AI Era > Source: > Published: 2026-07-08 13:00:00+00:00 # Why Unemployment Insurance Alone Is Not Enough in the AI Era ### Key Takeaways - Predictions about the integration of [artificial intelligence in U.S. workplaces](https://equitablegrowth.org/what-impact-is-artificial-intelligence-having-on-the-u-s-labor-market-and-the-nations-economy/)and how it might create disruptions in the U.S. labor market have increased attention on the nation’s[Unemployment Insurance](https://equitablegrowth.org/unemployment-insurance-reform-primer/)system. - The UI system was designed to temporarily offset workers’ lost income due to no-fault job loss and also act as a broader automatic stabilizer during economic downturns. The UI system was not designed to support the U.S. economy through major sectoral or structural labor market transitions. - Historically, the adoption of new, general-purpose technologies such as AI have changed the nature of work and led to job displacements during technological transitions. **What this means for growth:** Policymakers can support shared economic prosperity and growth by equipping the UI system to be responsive to already-familiar challenges in the U.S. labor market, supporting household consumption when workers suffer from involuntary unemployment, and enabling unemployed workers to engage in productive job searches or job retraining to secure new employment with higher pay. But potentially sweeping jobs displacements—if they do occur in the near future due to AI—will require other policy responses. ### Overview Since the large language model ChatGPT was rolled out in November 2022, [artificial intelligence](https://equitablegrowth.org/research-paper/navigating-the-research-on-the-impacts-of-ai-on-work-workers-and-the-labor-market/) has captured the public’s attention and generated myriad debates over the emerging technology’s potential impacts on the U.S. economy and labor market. Yet the role that artificial intelligence will actually play in the U.S. labor market and the economy in the coming years is still uncertain, though researchers are rushing to fill the void and [investigate the impacts that AI is having on work, workers, and the labor market](https://equitablegrowth.org/research-paper/navigating-the-research-on-the-impacts-of-ai-on-work-workers-and-the-labor-market/). Despite limited and mixed evidence on AI’s effects on workers’ productivity, headline news increasingly features [bold claims](https://www.axios.com/2025/05/28/ai-jobs-white-collar-unemployment-anthropic) that AI will lead to widespread unemployment as employers increasingly replace labor with advanced technologies. Whatever the future holds, policymakers must prepare for a range of scenarios that account for ways in which AI could evolve and become adopted in U.S. workplaces. Predictions about how the integration of [artificial intelligence in the workplace](https://equitablegrowth.org/what-impact-is-artificial-intelligence-having-on-the-u-s-labor-market-and-the-nations-economy/) might create U.S. labor market disruptions have increased attention on [Unemployment Insurance](https://equitablegrowth.org/unemployment-insurance-reform-primer/)—the [joint federal-state program](https://www.congress.gov/crs-product/IF10336) that provides temporary and partial wage replacement, or income support, to workers who lose their jobs through no fault of their own—as a potential policy response. But this raises the question: Is the nation’s UI system up to the task of supporting millions of U.S. workers in the event of widespread job dislocation? Unemployment Insurance was never meant to respond to large, prolonged labor market disruptions. The UI system also was not designed to be an anti-poverty program or a reskilling or retraining program. And, given its origins in the 1930s, it certainly was not designed with the widespread adoption of artificial intelligence technologies that could change the nature of work in mind. This does not diminish the role of Unemployment Insurance as a valuable income support program. Indeed, [changes in its delivery to eligible workers are important to enact](https://equitablegrowth.org/unemployment-insurance-reform-primer/), including increasing the duration of benefits, the level at which the program replaces wages, and easing the eligibility requirements. Unemployment Insurance is important to individual workers and their households because the benefits help to smooth [consumption](https://www.congress.gov/crs-product/IF11253) when they stop receiving paychecks. This, in turn, benefits the wider economy because it reduces financial volatility and helps to keep economic activity moving. Unemployment Insurance provides valuable income that allows laid-off workers time to find jobs that best match their skills rather than taking the first available jobs. This added time is particularly needed when [job searches take longer than they have in recent years](https://www.bls.gov/news.release/empsit.t12.htm). Workers benefit from better matches through higher wages and job satisfaction, and the wider economy benefits from the efficient allocation of labor. Indeed, [research](https://equitablegrowth.org/expanding-eligibility-for-unemployment-insurance-helps-low-income-u-s-workers-find-better-jobs/) finds that UI benefits can help low-income workers secure jobs that are more stable and higher paying than the ones they previously were in. [Other research likewise finds](https://equitablegrowth.org/working-papers/do-unemployment-insurance-benefits-improve-match-quality-evidence-from-recent-u-s-recessions/) that extending UI benefits can improve the quality of job matches for workers, particularly for women, non-White, and less-educated workers. While future labor market dynamics are uncertain, there are evidence-based ways to strengthen the UI system to be as resilient and responsive as possible, even if AI is not the principal cause of future economic upheavals. This brief outlines issues with the nation’s UI system that make it unprepared to handle a sudden and overwhelming tide of unemployed workers [absent congressional action](https://www.congress.gov/crs-product/R46687). It also details the limits of existing policies and programs, such as Unemployment Insurance, in addressing structural transformations of the labor market at a massive scale. ### Unemployment Insurance is not currently equipped for broader labor market transitions Unemployment Insurance is intended to support the U.S. labor market. The program’s benefits are tied to contributions based on wages to insure against the risk of wage losses. Participation is limited by eligibility, available funding, and beneficiaries’ abilities to navigate the administrative requirements of enrollment. Unemployment Insurance serves as both a partial wage replacement for involuntarily unemployed workers and as an [economic stabilizer](https://equitablegrowth.org/unemployment-insurance-reform-primer/) during downturns by temporarily injecting funds into the economy when workers are struggling to find employment due to broader macroeconomic conditions. The UI system, as it is, struggles to fulfill its designed intent. [Analysis of 2026 data](https://www.nelp.org/ui-benefits-have-not-kept-up-with-the-rise-in-unemployment/) shows that unemployment benefit claims are not increasing with known increases in the unemployment rate, revealing a disconnect between the UI program’s responsiveness and real job losses. Researchers point to the disparate rules and regulations that govern each state’s program as driving factors in this gap. The disconnect between UI claim rates and the unemployment rate has been documented for decades. In 2000, [the U.S. Bureau of Labor Statistics detailed why eligible jobless workers were not applying for benefits.](https://www.bls.gov/opub/mlr/2000/06/art2full.pdf) More than 20 years later, in 2022, [the agency reported](https://www.bls.gov/opub/ted/2023/most-unemployed-people-in-2022-did-not-apply-for-unemployment-insurance-benefits.htm) that almost 75 percent of potentially eligible unemployed individuals did not apply for unemployment benefits. Research finds that [eligible workers do not apply for benefits](https://www.nber.org/papers/w30266) for reasons including [confusion about](https://www.minneapolisfed.org/research/community-development-working-papers/understanding-eligibility-and-take-up-in-state-unemployment-insurance-programs) [eligibility](https://equitablegrowth.org/working-papers/barriers-to-benefits-unemployment-insurance-take-up-and-labor-market-effects/) rules, which results in fewer workers qualifying for benefits because of the changing [nature of work](https://documents1.worldbank.org/curated/en/816281518818814423/2019-WDR-Report.pdf), such as the rise in gig work, and a complicated application process, which deters even eligible people from applying for benefits. This gap between workers eligible for Unemployment Insurance and those who apply and receive UI benefits is among the reasons why the UI system is not prepared to support workers displaced by broad structural changes to the labor market, such as those posed by the advent of artificial intelligence. Other factors include the level of wage replacement that Unemployment Insurance offers to workers and the amount of time that workers are allowed to claim UI benefits. Let’s consider each in turn in the context of widespread adoption of artificial intelligence in U.S. workplaces. #### Eligibility for UI benefits Not all workers who could be affected by AI-related job disruptions will qualify for Unemployment Insurance. To receive UI benefits, workers must meet earnings thresholds distributed over a long enough period in their work records—[generally 12 months](https://www.congress.gov/crs-product/R48447)—and be actively searching for work. States primarily fund UI benefits via payroll taxes within a federal program framework, but states have flexibility to determine the length of time that workers receive benefits, the wage-replacement rate they receive for the duration of their benefits, and the platform through which workers can access benefits. UI systems thus exclude many entry-level workers or others re-entering the labor market, including [young college graduates](https://libertystreeteconomics.newyorkfed.org/2026/06/remote-work-leaves-younger-workers-sidelined/) who have been increasingly struggling to find employment and part-time workers who often lack sufficient earnings history to qualify for UI benefits. Research finds that workers who are unable to enter or re-enter the labor market suffer long-term consequences, such as [reduced earnings](https://www.aeaweb.org/articles?id=10.1257%2Fapp.4.1.1) and [slower career progression](https://www.sciencedirect.com/science/article/abs/pii/S0927537109001018) relative to peers who got jobs in healthier labor market conditions. While there is [no](https://libertystreeteconomics.newyorkfed.org/2026/06/remote-work-leaves-younger-workers-sidelined/) [conclusive](https://www.nber.org/papers/w33777) [research](https://equitablegrowth.org/research-paper/navigating-the-research-on-the-impacts-of-ai-on-work-workers-and-the-labor-market/) yet regarding how AI is affecting entry-level employment, the exclusion of these workers from most UI benefits is well-documented. Similarly, on-demand or “gig,” freelance workers, and self-employed workers also face barriers to accessing Unemployment Insurance. These workers are typically excluded from state UI systems because they are funded through payroll taxes paid by traditional employers, which this group of workers, typically [classified as independent contractors](https://www.epi.org/publication/gig-worker-survey/), tends not to have. On-demand, gig, freelance, and self-employed workers are no less exposed to the integration of AI into the labor market than workers in traditional employment arrangements. Differences in state rules also shape who has access to UI benefits. Research finds that, in practice, there are stark racial disparities in who receives Unemployment Insurance. States with more Black workers have more stringent rules overall, and Black claimants have an [8.4 percent lower wage-replacement rate](https://equitablegrowth.org/new-research-shows-that-differences-in-state-unemployment-insurance-rules-cause-inequity-and-inefficiency/) than White claimants. Even during [periods of time when UI benefits were expanded](https://equitablegrowth.org/working-papers/understanding-disparities-in-unemployment-insurance-recipiency/), such as during the COVID-19 pandemic of 2020–2023, less-educated workers, younger workers, and non-White workers reported lower rates of accessing UI benefits. #### UI wage-replacement levels For Unemployment Insurance to serve its dual purpose of partially replacing lost wages and boosting the economy during downturns, the wage-replacement rate must be generous enough to support households’ needs during spells of unemployment. By law, states are allowed to determine the wage-replacement rate they offer to unemployed workers. The replacement rate varies widely by state, despite geographic drivers of different costs of living in each state already being accounted for in workers’ underlying earned wage histories. Some states, including Florida and Arizona, replace less than 15 percent of average weekly wages. These low replacement rates have led [federal policymakers](https://equitablegrowth.org/an-unemployment-insurance-modernization-bill-now-before-the-u-s-senate-is-a-much-needed-step-in-the-right-direction/) and [experts](https://equitablegrowth.org/unemployment-insurance-reform-primer/) to propose national replacement rates for workers closer to 50 percent or 75 percent of workers’ wages. After the [Great Recession](https://equitablegrowth.org/unemployment-insurance-reform-primer/) of 2007–2009, many states reduced their wage-replacement rates due to diminished UI funding. Some researchers believe the impacts of AI on the U.S. labor market may vary geographically. If the United States were to experience concentrated job losses in places with particularly low UI benefits, then the ripple effects on the wider economy could be larger and longer-lasting, as observed [in North Carolina](https://www.nber.org/papers/w30152), which saw declines in the quality of workers’ job matches and wages following cuts to UI benefit amounts and duration in 2013. Although some opponents of raising the wage-replacement level argue it will serve as a disincentive for workers to return to work, the duration and amount of UI payments to workers at their current levels have a limited impact discouraging workers from engaging in the labor market. [Research using administrative data from California between 2000 to 2020](https://www.capolicylab.org/wp-content/uploads/2024/01/UI-Benefit-Generosity-and-Labor-Supply-from-2022-to-2020-Evidence-from-California-UI-Records.pdf) finds that UI benefits had little negative distortionary effects on labor supply in the state. #### Length of UI benefits The length of time unemployed workers can claim UI benefits is tied to the program’s goal of supporting workers between jobs. It was not intended to serve as a retraining or reskilling support program, although it was once more closely linked with the [Employment Service](https://www.dol.gov/agencies/eta/performance/results/wagner-peyser), a [New Deal program](https://research.upjohn.org/reports/29/) that provided job seekers with resources such as job matching and counseling and supported employers by referring suitable job candidates for vacancies. Policymakers [could foster this link once again](https://research.upjohn.org/up_workingpapers/269/). States determine how long unemployed workers can receive unemployment benefits. [Most states](https://www.congress.gov/crs-product/IF10336) provide UI benefits for up to 26 weeks. Some states, including Florida, Kentucky, and North Carolina, only provide benefits for a maximum of 12 weeks, while others, such as Massachusetts, provide up to 30 weeks. Economists agree that during recessions, people tend to be unemployed for longer spells of time and the UI system should accommodate that by offering benefits for longer periods. One solution is the federal-state [Extended Benefits program](https://equitablegrowth.org/an-unemployment-insurance-modernization-bill-now-before-the-u-s-senate-is-a-much-needed-step-in-the-right-direction/), which offers an additional 13 weeks to 20 weeks of UI benefits to workers in states with high unemployment rates. First passed in 1970, the Extended Benefit system has played an exceedingly limited role in supporting workers during recessions due to flawed trigger requirements to enable expanded benefits. Instead, [Congress has acted nine times](https://www.congress.gov/crs-product/IF10336) to temporarily expand Unemployment Insurance in response to recessions so as to increase UI benefits in response to economic recessions or crises, most recently during the COVID-19 pandemic. Under conventional conditions, UI benefits are not sufficient to support the basic needs of low- and middle-income workers during periods of unemployment without eating into their personal savings—which, it should be noted, most American households lack. According to [research](https://equitablegrowth.org/working-papers/income-after-loss-of-extended-unemployment-benefits/) covering data from the dot-com recession of 2001 and the Great Recession in 2007–2009, many UI recipients did not find new jobs before exhausting their benefits, even when benefits were extended by Congress during these two recessions. Researchers found that once households exhausted their UI benefits, they suffered dramatic reductions in household income and consumption unless they were able to transition to other social programs such as the Temporary Assistance for Needy Families program or the Supplemental Nutrition Assistance Program. The length of time workers are able to receive UI benefits is particularly salient to artificial intelligence debates. The potential labor market disruptions from AI include the possibility that workers with similar skillsets, such as [logistics operation workers](https://equitablegrowth.org/adoption-of-generative-ai-will-have-different-effects-across-jobs-in-the-u-s-logistics-workforce/), could become unemployed at the same time across multiple sectors and may face higher competition and thus require more time to find new employment. Such a situation would likely necessitate deploying [evidence-based](https://www.brookings.edu/articles/ai-labor-displacement-and-the-limits-of-worker-retraining/) worker-transition policies, including worker retraining and reskilling programs to support workers as they reenter the workforce. ### Policymaker considerations The nation’s unemployment system requires attention from federal policymakers. Advances in artificial intelligence have underscored this need as employers point to AI as the driving force behind increased layoffs, regardless of evidence supporting such claims. Policymakers should expand and reform the UI system to more appropriately meet the next moment of economic contraction. Yet the UI system would likely still be unprepared to address the needs of workers in the event of widespread AI-related job displacements, given the program is not a panacea for dealing with labor market disruptions. In short, Unemployment Insurance is not designed to address human capital mismatches, occupational segregation, occupational exposure to AI, or the potential spillover effects that job losses and lost wages can have on the wider economy. The UI system was built to support individual workers and cyclical shocks, not the widespread restructuring of the labor market itself. The program only works as a stabilizer to a certain extent. When there is a massive economic disruption or recession, it will fall short, which is why Congress historically has had to step in to temporarily expand benefits in response to macroeconomic conditions. [Proposals abound](https://equitablegrowth.org/how-to-stop-a-recession-by-strengthening-income-supports-in-the-united-states/) to improve the UI program’s ability to spring into action as an automatic [stabilizer](https://www.hamiltonproject.org/publication/policy-proposal/unemployment-insurance-and-macroeconomic-stabilization/) in the case of widespread labor market disruptions. These reforms are necessary to keep the UI system healthy and responsive. But even these proposals would fall short of what would be needed if sectoral or structural changes lead to broader job losses driven by AI or other forces. Policymakers must consider the limits of existing policies and programs, such as Unemployment Insurance, in addressing structural transformations of the labor market at a massive scale. They can and should look to [past government programs for lessons learned](https://equitablegrowth.org/research-paper/lessons-from-past-trade-adjustment-policies-to-support-displaced-workers-in-the-era-of-artificial-intelligence/) and should consider other levers available to them as they confront the AI challenge to meet the needs of both traditional and nontraditional workers, including on-demand or gig workers, entry-level workers, part-time workers, and immigrant workers. Future economic policy must ensure AI-related workplace disruptions do not stratify the U.S. workforce into those with and those without the resources to adapt and thrive in an AI-dominant economy. While we wait for the future to arrive, policymakers can support shared economic prosperity and growth by equipping the UI system to be responsive to the U.S. labor market, supporting household consumption when workers suffer from involuntary unemployment, and enabling unemployed workers to engage in productive job searches or job retraining to secure new employment with higher pay. *Did you find this content informative and engaging?*[Get updates](https://equitablegrowth.org/engage/get-updates/) and stay in tune with U.S. economic inequality and growth!