There is a seductive idea gaining traction in regulatory circles: that emerging technology sectors can largely police themselves, with light-touch government oversight serving mainly as a backstop. We are told this approach is pragmatic, inevitable, and the only way to avoid stifling innovation.
It is being sold as inevitable. It deserves more skepticism than it is getting.
The self-regulation narrative appears across multiple domains right now. Crypto advocates argue that industry standards and best practices can prevent fraud without heavy-handed legislation. AI safety proponents suggest that company ethics boards and voluntary safety protocols can manage risks. Tech companies consistently pitch internal governance as superior to external regulation, arguing they understand their systems better than any government body could.
There is a surface logic here. Industries do move faster than bureaucracies. Companies do have strong incentives to avoid catastrophic failures. Self-regulation can be more flexible and adaptive than statutory rules written years before technology matures.
But the historical record offers consistent cautions that deserve consideration before accepting self-regulation as the default mode.
The financial services industry operated under a voluntary compliance framework for decades before 2008. Banks had internal risk management committees, industry associations published guidelines, and executives claimed robust self-policing. The result was not optimal. When incentives misaligned with broader stability, self-regulation proved insufficient. This is not controversial history; it is documented extensively in post-crisis reviews.
The pharmaceutical industry similarly relied on substantial self-regulation for drug testing and safety reporting before and after market approval. Over time, patterns emerged: companies prioritized speed and profitability when self-monitoring, and systematic problems emerged only when external pressure mounted.
None of this means self-regulation is worthless or that government agencies are automatically superior. The point is different: voluntary frameworks tend to work adequately when incentives align with public interest, and tend to erode when they diverge.
Today's emerging tech sectors present exactly the conditions where incentive divergence is likely. Crypto platforms profit from trading volume and user growth, not from preventing fraud or systemic risk. AI companies benefit from rapid deployment, not from exhaustive safety testing. Social media platforms monetize engagement, not truthfulness. These are not character flaws; they are structural features of how these markets work.
Self-regulation enthusiasts respond that reputational damage and user exodus provide market discipline. Perhaps. But "perhaps" is not the same as "certainly," and the cost of being wrong about emerging financial or safety-critical systems can be substantial.
This does not mean statutory regulation is obviously the right answer either. Badly designed rules can calcify around outdated assumptions. They can cement incumbent advantages and discourage experimentation. Regulators often lack technical expertise. The regulatory capture problem is real.
The reasonable position sits between extremes: assume that self-regulation will be incomplete, that external oversight serves a genuine function, and that the appropriate balance depends on specific risk profiles. A platform for trading digital assets presents different risks than an experimental AI system, which presents different risks than a social network.
What warrants skepticism is not regulation itself, but the framing that self-regulation is either inevitable or sufficient. That framing often reflects the preferences of regulated industries rather than hard analysis of what actually works.
Industries will continue advocating for light oversight. That is their job. Policymakers should listen to their input, but not accept their premise about inevitability. The evidence suggests that some combination of industry standards and external accountability tends to produce better outcomes than either alone.
That is not a flashy message. It is messier than "let markets decide" or "regulate everything." But messy realism beats confident inevitability.