This piece was originally published in Asian Labor Futures, a Substack by 2024 Just Tech Fellow Kriangsak Teerakowitkajorn. We have republished it here with permission from the author with minor edits. For more insights into technology and labor in Asia, subscribe to Asian Labor Futures.

Labor’s Groundhog Day

Daily AI development may seem like a new phenomenon, but the discussions surrounding workers in AI supply chains feel eerily familiar. The terms “ethical” or “responsible” AI, along with conversations about AI governance and accountability, echo the longstanding battles against labor exploitation in the global supply chains.

Corporations have historically engaged with voluntary codes of conduct to avoid enacting meaningful change. Once again, workers are being exploited while governments across Asia draft new AI laws riddled with buzzwords like competitiveness, sovereignty, and risk.

Let’s get to the heart of the matter: The core issue remains the same as always—to ensure that fundamental labor rights are upheld as nonnegotiable!

Unfortunately, in many parts of Asia, minimum wages have effectively become lifelong wages, and core labor rights are still largely unattainable. AI supply chains—the interconnected global processes that create “artificial intelligence,” from extracting raw materials to manufacturing hardware and training AI models— do not deviate from this trend; instead, they perpetuate it, reinforcing the exploitation of workers while concentrating wealth at the top.

In the Philippines, business process outsourcing call center and data workers recently demanded the resignation of Maria Criselda Sy, the executive director of the National Wages and Productivity Commission, after she dismissed them as “just minimum wage earners,” blaming their struggles on a lack of education. Workers pushed back:

“The real issue here is low wages, contract work, regional wage boards, and a rotten system designed to make Filipino labor cheap and easy to exploit.”

Labor Is Nowhere to be Found

The Tech Global Institute’s Labor in the Shadows report—one of the first comprehensive studies of data work with a focus on Asia—reminds us of what should already be obvious: AI is built on the backs of invisible workers. India and the Philippines are central to this hidden workforce, employing millions through business process outsourcing firms that manage tasks ranging from content moderation to data annotation.^[1]Disha Verma, Labor in the Shadows: Rights and Risks for Asia’s Data Workers (Tech Global Institute, 2025), 1–2, … Continue reading Behind the narratives of digital innovation, these workers endure low pay, extreme job insecurity, and minimal protection.

One moderator in Hyderabad described their job as entering a “torture chamber” each day, while workers in Manila report experiencing nightmares, emotional breakdowns, and even suicide attempts. When psychological services are provided, they mainly serve as compliance tool for the companies.^[2]Verna, Labor in the Shadows, 16.

The report also reveals that exploitation is systematically engineered into the digital workplace. Job roles are often camouflaged with vague titles like “process executive” or “partner,” stripping workers of their rights. Layers of subcontractors obscure accountability, nondisclosure agreements (NDAs) and algorithmic monitoring silence workers, and blacklisting punishes dissent harshly. Companies can terminate contracts at will, relocate operations across borders, or replace human labor with AI without any obligation to the workers.

Despite these pressing issues, governments across Asia continue to draft AI and digital economy laws as if these workers did not exist. For global companies, Asia represents precisely what they seek: a large pool of inexpensive, compliant labor, and governments eager to promote “competitiveness,” all while ignoring the issues of wages, safety, and dignity.

From my own years of engaging national and regional governments through the Just Economy and Labor Institute, this reality is all too familiar: Technology is consistently treated as if it were separate from labor. After more than a decade of advocating for workers’ rights—including alongside industrial workers employed by multinational corporations—the struggles of those who endured workplace violations, compounded by official neglect, remain too vivid to forget.

What I have come to see is that, in our context, regulation means something entirely different: It aligns the interests of states and corporations, openly dismissing workers’ concerns about precarity, health, and safety, and enforces compliance through disciplinary measures.

Cutting through the Noise

Working alongside civil society and labor organizations for years has taught me one important lesson: Progress in Asia only comes from mass mobilization and pressure from the bottom up.

Policymakers and politicians become accountable only when workers organize and act collectively on the streets—not when experts and officials share stories of workers’ struggles at international forums.

In the book Feeding the Machine: The Hidden Human Labor Powering AI, Callum Cant, James Muldoon, and Mark Graham outline five concrete steps toward a world in which all AI workers are treated with dignity and respect:^[3]Callum Cant, James Muldoon, and Mark Graham, “Reviewing the Machine,” chap. 8 in Feeding the Machine: The Hidden Human Labor Powering AI (Bloomsbury Publishing, 2024).

1. Building the collective power of worker organizations.
2. Civil society organizing to hold companies accountable.
3. Enacting stricter regulations on AI companies.
4. Establishing forms of worker governance and ownership over enterprises.
5. Challenging the injustices of the broader system within which these companies operate.

They argue that these steps can be mutually reinforcing: Stronger regulations empower workers in negotiations; worker ownership enhances accountability; civil society support can amplify worker struggles.

While I agree with this general argument, I see added complexities for us in Asia. Given the current hype surrounding AI and the trajectory of platform and data worker activism over the past decade, the real tension lies between the first and second strategies: building the collective power of worker organizations and organizing civil society to hold companies accountable.

Too often, civil society efforts—centered on name-and-shame strategies led either by consumer campaigns or NGO-led accountability reports—end up validating corporate delay strategies while reinforcing governmental neglect in lawmaking. From my experiences, this dynamic frequently pits one strategy against another, weakening collective pressure on corporations to engage directly with worker groups.

In such scenarios, lawyers and consultants thrive, while workers are further marginalized. Unless campaign strategies are grounded in worker struggles, we risk replaying the cycle of “ethical” audits and voluntary codes of conduct that never shift power.

Footnotes