
...


By Matthias Oschinski and Mina Narayanan
0
Share
Newsweek is a Trust Project member
See more of our trusted coverage when you search.Prefer Newsweek on Googleto see more of our trusted coverage when you search.
An AI system passes a medical exam. Another writes working software in seconds. A third generates a photorealistic face that never existed. And every time, the same debate follows: Should we let this technology run or rein it in?
Innovation versus regulation. Speed versus safeguards. Move fast or fall behind. These binaries dominate every serious policy conversation about AI, industrial strategy and climate. They also happen to be the wrong frame entirely. The real question is not whether to innovate or regulate. It is: What are we innovating for—and for whom?
Innovation is a means, not an end in itself. Its value depends entirely on the outcomes it produces. Almost everyone agrees on what those outcomes should be: rising living standards, durable economic growth and workers who share in the gains of technological change, rather than bearing the costs. The problem is not that regulation hampers innovation. The problem is that innovation without purpose is a dead end.
Read More on Opinion
Nowhere is this more urgent than in the labor market. Four decades of rising inequality and a declining labor share have already demonstrated what happens when technological change leaves workers behind. AI will either accelerate those trends or reverse them. And the design choices being made right now will be difficult to redirect later.

...
Consider the difference between two uses of the same large language model. One automates customer service roles, cuts headcount and sends the savings to shareholders. The other gives front-line employees a tool that handles routine queries so they can focus on complex problems, improving job quality and customer outcomes alike.
This is not a hypothetical distinction. A growing body of research makes clear that the critical divide in AI is not between technologies that help and technologies that harm. Rather, it is between systems that commodify human expertise and those that extend it, enabling workers to tackle new problems, exercise deeper judgment and acquire skills they could not develop alone. When AI is designed to collaborate with workers rather than replace them, it can act as a force multiplier for human capability. But that design does not happen by default.
Left to market incentives alone, AI development tilts toward automation. The firms building these systems profit from scale and cost reduction, not from strengthening worker capabilities. Their business model rewards tools that make expertise disposable. And a pervasive fixation on artificial general intelligence—systems that outperform humans at everything—actively crowds out investment in AI that works with people rather than instead of them.
To some extent, this fixation absolves firms of demonstrating how system capability improvements in the lab actually translate into shared prosperity or better jobs, potentially leading to a gap between new technologies and the problems they purport to address. Therefore, the absence of governance is far from neutral—it shapes innovation in often socially suboptimal ways that are disconnected from real-world problems.
This is precisely why governance matters. If innovation’s direction is shaped by incentives and today’s incentives favor automation over augmentation, then the question is not whether to intervene but how to intervene most effectively. Well-designed governance does not suppress innovation. Instead, it shapes the direction of innovation in socially beneficial ways by prioritizing economic resilience, higher living standards and worker empowerment.

...
Critics of governance may counter that the act of governing is vulnerable to regulatory capture and accessible to only a privileged few. While this is a legitimate concern, governance encompasses many different mechanisms—from Congressional hearings to agency requests for information—that provide opportunities for diverse stakeholders to shape the outcomes of policymaking. By contrast, surrendering goal-setting to a narrow set of private interests forecloses this type of public participation.
Critics also argue that regulation imposes burdensome compliance costs and paralyzes firms’ ability to innovate. It is true that poorly designed regulation may fail to achieve its intended goals and unfairly penalize companies, especially small and medium-sized companies. However, thoughtfully designed governance should support—and not stymie—the design and adoption of innovations that advance shared goals.
What does thoughtfully designed governance for AI look like? It may entail evaluating AI capabilities to inform the development of safety features, building institutional capacity to enforce compliance with AI rules, or making resources available to less-established enterprises and researchers who invest in AI safety and alternative development models. Well-designed AI governance may also constitute on-the-job training that prepares workers for an increasingly AI-augmented world.
More generally, AI governance reduces uncertainty about liability, standards and acceptable use, which is itself a strategic risk in high-stakes technologies. When rules are unclear, incentives favor short-term deployment over robustness and safety, which may impact AI adoption over the long run. Well-designed governance ultimately enables investment in safety, evaluation and long-term deployment, leveling the playing field for firms to compete on trustworthiness and performance.
The real debate is not innovation versus regulation. It is innovation guided by purpose versus innovation left to default incentives. Countries that align innovation with shared goals—rising living standards, economic resilience, a workforce that grows with technology rather than being discarded by it—are more likely to sustain public trust, attract long-term investment and build competitive advantage that lasts.
This requires more than good intentions. It requires policymakers who treat governance as infrastructure, not obstruction. It requires firms willing to invest in tools that extend human capability, not just cut costs. And it requires a public conversation that moves past the false binary of innovation versus regulation and toward the harder, more productive question of what we are innovating for. That question is not a brake on progress. It is how progress becomes meaningful.
Matthias Oschinski is a senior fellow at Georgetown University’s Center for Security and Emerging Technology (CSET). Mina Narayanan is a research analyst at Georgetown University’s Center for Security and Emerging Technology (CSET).
The views expressed in this article are the writers' own.
View 0 Comments
Submit Correction
View Editorial & AI Guidelines