In the Face-Off Against AI, Funders Need to Have Workers’ Backs

The call is coming from inside the house. The CEO of the artificial intelligence company Anthropic recently issued a stark warning: AI, he said, could obliterate 50 percent of all entry-level white collar jobs by 2030 and drive unemployment rates as high as 20 percent. His prediction underscored the massive scale of AI disruption and the urgent need for action, including from philanthropy.

This isn’t a problem for the distant future. AI is already transforming the American workplace — enabling greater efficiencies and productivity but also leading to job displacement. The unemployment rate for recent college graduates is now higher than the national average, in part because employers are using AI instead of hiring entry-level workers.

Even if people keep their jobs, the rapid adoption of AI threatens their power and agency in the workplace. Many workers are training AI systems that could eventually de-skill, displace, or surveil them — often without their having any say in how the technology is developed or deployed.

The pressure on employees is slowly mounting. But a critical window for intervention remains open as policymakers and companies hash out AI governance practices. For philanthropy, this moment demands more than isolated investments in reskilling workers displaced by new technology. It calls for a coordinated strategy that ensures workers have a voice in shaping the AI-driven future.

An AI Power Struggle

Grant makers frame AI as a technological issue, but it’s fundamentally about power. Beyond simply asking what AI can do, they need to determine who decides how it’s used and who benefits.

Currently, these systems are primarily being developed by and for corporations. Affected workers have little input, although they understand that the AI revolution isn’t happening with them but to them. Polling by the organization Mike leads, the Omidyar Network, found that most workers expect AI’s productivity gains to mostly benefit employers.

Moreover, only 39 percent of unionized workers and 34 percent of nonunionized employees believe they can control the future of digital technology in their workplaces. This power imbalance affects low-wage workers the most. They have little say in workplace technology decisions and are more susceptible to AI-enabled surveillance and automation.

Giving workers a voice in AI development won’t stop technological progress but will ensure technology advances wealth and opportunity for everyone — not just shareholders and executives.

Combining Causes

Philanthropy has long funded both AI research and development and worker organizing. Now it needs to integrate AI into its labor rights, equity, and racial and economic justice initiatives.

Grant makers might assume that unions alone can take on this work. Some, in fact, have made important strides. During the 2023 Hollywood writers’ strike, for example, the Writers Guild of America and SAG-AFTRA unions won protections for writers against film and television studios’ use of AI. And the California Federation of Labor Unions is sponsoring three bills that limit AI surveillance.

But only 10 percent of American workers belong to unions, and most threatened jobs are in nonunionized sectors. Organized labor can’t tackle the job alone. Philanthropy has capital, connections, and convening power that make it uniquely positioned to sound the alarm and support new solutions. Here’s a three-pronged approach that a wide range of funders could embrace.

Fund worker-led, grassroots organizing. With philanthropic support, unions and worker-focused nonprofits can channel their AI anxiety into organizing efforts and advocate for tech laws and corporate policies that reflect worker needs.

For example, with support from the Omidyar Network and the Ford Foundation, the nonprofit Model Alliance helped pass New York State’s Fashion Workers Act, which gives models control over digital replicas of themselves, among other protections. The first law in the nation regulating generative AI’s use in the workplace, it offers a blueprint for how philanthropy can help workers win AI protections through grassroots advocacy. Similar efforts are underway to support gig drivers and truckers.

At a minimum, philanthropy should help workers hone their AI literacy by supporting initiatives that connect them with tech experts. The James Irvine Foundation, which Don leads, and Freedom Together, along with the Omidyar Network and Ford, co-funded a statewide meeting in California earlier this year that brought together nearly 300 people from unions, worker advocacy organizations, community groups, and tech companies to strategize how to protect workers from AI’s potential harms. More such cross-sector conversations will help workers strengthen their ability to shape AI on their terms.

Advocate for employees from the top down. Beyond grassroots organizing, grant makers can also represent workers’ needs in policy discussions and corporate decision-making processes where grant makers may have a seat at the table but workers typically don’t. That could involve supporting the creation of industrywide standards for AI usage, safety certifications, and accountability mechanisms that ensure fair, transparent, and equitable AI use.

For instance, Pennsylvania’s Generative AI Pilot Program brought together public sector leaders, OpenAI, academics, and the local Service Employees International Union chapter to determine how to responsibly integrate AI into government operations. The program incorporated employee feedback throughout implementation and established a worker oversight board. Ultimately, it improved employee efficiency while ensuring that AI assisted rather than replaced workers. Philanthropy should support similar inclusive processes that could serve as a model for other states and sectors.

Philanthropies that have an investment arm — as the Omidyar Network does — can also back companies that responsibly create and deploy technology and encourage them to incorporate employee needs into AI design. Workplaces that include employees early and often in AI decision making are more likely to weather the coming storms.

Help workers develop AI tools that meet their needs. Philanthropy can fund projects in which workers co-design AI systems for their own benefit and pilot programs that demonstrate why such tools are useful. With support from the James Irvine Foundation, the National Domestic Workers Alliance developed the La Alianza chatbot to share news and resources with Spanish-speaking domestic workers. The chatbot also sends surveys to workers about the challenges they face, which then inform the National Domestic Workers Alliance’s organizing efforts.

But without more funding, most workers won’t have access to such tools. In a recent Chronicle of Philanthropy survey, almost nine in 10 nonprofit leaders said budget constraints hamper tech adoption and less than half said they use AI tools.

The Omidyar Network and the Irvine, Ford, and Freedom Together foundations are among the grant makers that have begun integrating AI concerns into their economic and social justice funding. But more grant makers should join this effort, recognizing the connection between worker power and AI governance.

As Anthropic’s CEO made clear, the AI revolution is happening with or without worker input. A future is possible in which AI shrinks rather than widens the wealth gap and improves rather than degrades quality of life. By funding worker power in the age of AI, philanthropy can help rewrite the script before inequality gets baked into the code.