Job losses are starting to accelerate due to AI, robotics, and automation. Anthropic CEO Dario Amodei warns that AI could cut half of all entry-level white-collar jobs in the coming years. A McKinsey report finds that “current gen AI and other technologies have the potential to automate work activities that absorb up to 70 percentof employees’ time today.” A Wall Street Journal article also cites leading CEOs who claim “AI will wipe out jobs.” OpenAI CEO Sam Altman has ominously written that the “2030s are likely going to be wildly different from any time that has come before.”
AI’s impact on the workforce is hard to forecast because so many factors affect economic projections. There are uncertainties concerning trade imbalances, exchange rates, technology deployment rates, shifts in business models, political developments, geopolitical forces, and CEO proclivities toward their own workers. Former Harvard University President Larry Summers once quipped that predictions are so fraught with error that prognosticators never should use a date and prediction in the same sentence.
But given the magnitude of the phenomenon, it is important to think about ways to help workers who are likely to bear the impact of AI-related job losses. As I argue in my Brookings Press book, “The Future of Work: Robots, AI, and Automation,” several policy reforms would help people navigate the transition to a digital economy. This piece offers an overview of them.
Encourage companies to retrain workers
Businesses are on the frontlines of worker layoffs and have some responsibility for retraining employees. They need to finance job reskilling and upskilling for their employees so those individuals don’t get left behind. Governments should consider expanding tax credits for businesses that retrain laid-off employees, as these investments benefit the entire society. It is not in the country’s national interest for large numbers of people to become part of a permanent underclass struggling and failing to pay their monthly expenses despite their best efforts. Replacing these workers with AI would doom many individuals to a dismal financial future and create a host of social, economic, and political problems, many of them unique to AI’s role in worker displacement.
Make health benefits portable
It is vital that workers retain access to health insurance if they lose their jobs. Since much of American healthcare is handled through employers, job-related disruptions are especially harmful for workers. They may lose access to insurance or be forced to shift to another medical network that does not include their current healthcare providers. One of the benefits of the current Affordable Care Act exchanges is healthcare coverage offerings for those between jobs. Given the increasing threat of AI-related job losses, policymakers should ensure these mechanisms remain strong in safeguarding American workers’ access to healthcare as job disruptions spread.
Reduce vesting requirements for retirement benefits
Many organizations have lengthy vesting requirements, sometimes spanning three to six years, before people become eligible to get employer matches for retirement and keep that money. With the increasing rates of “job churn” due to AI and automation, people may end up working for many different organizations in a span of a few years. Lengthy vesting requirements will hurt workers who lose vesting time at each job transition, substantially jeopardizing their financial futures post-retirement. Employers should consider reducing vesting time to three or six months to aid those moving from job to job.
Loosen job licensing requirements
People who lose their jobs due to AI or automation may want to break into other occupations, but these shifts are often not easy. Many new occupations require specialized training, certification, and testing that can take months or sometimes years to secure. When certification is related to health and safety, these state-level licensing requirements make complete sense. Yet, there are some certifications and testing requirements not linked to health and safety that states should consider loosening to make it easier for people to enter new fields. This is especially true for individuals who are otherwise well-equipped to safely practice in those fields but face limitations due to their nontraditional education or language proficiency. For example, some states require degrees even if the individual has years of experience doing certain tasks. Policymakers could change those requirements to allow educational credentials or work experience to fulfill certification requirements.
Create worker retraining accounts
In the advent of AI, to incentivize society-wide upskilling, policymakers should consider creating worker retraining accounts that function similarly to retirement accounts. These accounts enable workers to use tax-deferred money to pay for job retraining. The program will help people gain new skills to keep up with the pace of innovation and navigate the vicissitudes of the digital marketplace. To sustain advancements in American innovation, policymakers should consider creative ways to maintain a digitally apt workforce.
Pay earned income tax credits monthly
Many individuals receive earned income tax credits each year after they file their income taxes. This allows them to get refunds if they have low incomes or high childcare expenses. Unless people want to make a large purchase or upgrade their homes, an annual payment will not help them with monthly expenses or provide a regular source of income. Making these payments monthly would create flexibility in terms of income support and provide more continuity over time, especially for workers in more vulnerable frontline jobs.
Clarify independent contractor rules
There should be clearer rules regarding the classification of workers as independent contractors. Right now, many individuals who work full-time are not classified as full-time employees and are thereby ineligible for health, retirement, or disability insurance benefits. This phenomenon is particularly prominent in the tech industry, which heavily relies on temp workers to perform core functions such as data analysis, content moderation, or administrative tasks. Having clearer rules would ensure that those working full-time get the benefits traditionally associated with full-time employment.
Fund job retraining programs in higher education
Some community colleges and other institutions of higher education are offering retraining programs for adults who need to upgrade their skills. As a way of assisting with the transition to a digital economy and helping individual workers suffering from job losses, governments should consider financially supporting these programs. Colleges and universities are a vital part of the retraining ecosystem and deserve to be helped as they carry out this important mission.
Make sure laid-off workers have access to high-speed internet
Twenty-four million Americans still lack access to high-speed internet at home, hindering their ability to utilize digital job boards and resources that help them learn new skills. For instance, many adult retraining programs take place online. To ensure all American workers can access these retraining programs, policymakers must guarantee all Americans have access to high-speed internet. My colleague Nicol Turner Lee has argued in her book, “Digitally Invisible: How the Internet Is Creating the New Underclass,” that a lack of access to high-speed broadband dooms people to poverty and creates huge problems for the individuals affected, as well as the country as a whole. As more job listings, training, and hybrid opportunities migrate online, it is imperative that individuals and employers reduce the barriers to seamless access and help those seeking new positions.
Ensure data-based job evaluations are fair
Employees are increasingly performing their tasks on computers, generating data analytics known as key performance indicators (KPI) that can serve as the basis for determining who gets laid off. As hybrid work becomes increasingly common, employees may receive laptops from their employers and take them home, which may inadvertently allow employers access to online activity that takes place on company equipment or networks. Individuals may be downgraded or lose their jobs because of data analytics that may portray them unfairly. Business leaders need to make sure these job-related analytics are fair and impartial and not unjustifiable intrusions into personal privacy and ensure that employees have clear expectations about their privacy in the hybrid workplace. Employers must also take into account those who successfully integrate AI into their job functions and judge them fairly for their performance rather than traditional productivity metrics.
Mitigate disparate job impacts
Not every worker is going to share the same workplace experiences. There are known disparities by race, gender, age, disability, immigration status, and veteran status, among other characteristics. Some individuals receive visas linked to a specific job or company, and encounter problems when they are laid off. Policymakers should make sure the impact of AI-related job losses does not fall disproportionately on those least able to retrain themselves, get new jobs, or obtain appropriate visas. Otherwise, there will be disparate impacts on marginalized populations that guarantee a poor economic future.
Consider a 4-day workweek as workers become more efficient and productive
As workers become more efficient and productive due to AI and other digital tools, companies should share the profits with them by considering four-day workweeks. Some employers have moved in this direction and have found positive results. One software company saw a 130% increase in revenue in addition to fewer sick days. Forward-looking companies should experiment with novel methods of work management and reward employees whose hard work and efficient performance improve the bottom line.
These and other recommendations require careful consideration by policymakers, industries, and employees, so that the AI transition can succeed with fewer casualties in the workplace and the broad labor market.