Conscious Policy for Unconscious Systems: Precaution, Power, and Transatlantic Drift in the Age of AI

Executive Synthesis – 2025

In 2001, I wrote that the starkest difference between Europe and the United States could be seen from the air: Paris ringed by farms, New York by sprawl. Underneath that visual was a deeper contrast—how each society approached uncertainty, risk, and public interest through policy. A quarter century later, the divergence remains—but the stakes have grown far stranger.

Europe’s approach to regulation has long been guided by the precautionary principle—a framework that allows policy to intervene in the face of scientific uncertainty to prevent potential harm. Whether in agriculture, pharmaceuticals, or environmental health, European law prioritizes a “better safe than sorry” ethos, even when risks are not fully understood. In contrast, the United States typically employs a risk-based model, requiring conclusive evidence of harm before regulatory action is taken. The burden of proof falls not on innovators, but on those harmed.

This transatlantic divergence, once a source of slow-moving WTO disputes over hormones in beef or GMOs in corn, has become the fault line for a new kind of policy battle—one that spans everything from artificial intelligence to food security and pandemic recovery.

From Purple Carrots to Neural Nets

The most visible battlefield today is artificial intelligence. In November 2024, an AI welfare officer at Anthropic, Kyle Fish, co-authored a report suggesting a non-negligible (15%) chance that current AI systems might already exhibit consciousness. It’s a provocative figure—widely debated and methodologically fuzzy—but it speaks to a broader concern: we do not fully understand how these systems work.

As Professor Murray Shanahan of Google DeepMind noted, even leading scientists admit we lack a reliable theory of how large language models (LLMs) achieve their surprising fluency. This opacity is not merely technical—it is philosophical, political, and moral. And it is precisely where the precautionary principle becomes relevant.

The EU’s Artificial Intelligence Act reflects this view, embedding ex ante controls on high-risk AI applications. It regulates the means of development—not just the harms after the fact. The United States, by contrast, continues to rely on voluntary guidelines and post-hoc liability, even as both parties in Congress push AI deregulation in the name of innovation and economic dominance.

This divide mirrors older debates about genetically modified organisms (GMOs). In the 1990s and early 2000s, the U.S. promoted biotech crops with minimal labeling requirements, while the EU, invoking the precautionary principle, required mandatory labeling and approval processes. American policymakers derided this as “anti-science.” European regulators called it democratic accountability.

Fast-forward to 2025: those same frameworks now shape responses to AI-generated deepfakes in elections, synthetic meat in food supply chains, and experimental neural implants in medical care. What once seemed a quaint European aversion to Monsanto now resonates in debates about OpenAI and brain-computer interfaces.

A Trade War of Worldviews

The current U.S.–EU trade tensions—ostensibly about agricultural subsidies, data flows, and carbon border taxes—are in fact philosophical. The Biden administration, despite its rhetoric of values-based diplomacy, continued to pressure Brussels to relax its precautionary framework in order to boost American exports and digital dominance. A new Republican Congress is doubling down on deregulation—going so far as to remove CO₂ from the list of EPA-regulated pollutants, a move seen by some as greenlighting fossil-powered AI supercomputers.

Critics argue that precaution is protectionism by another name—that it slows innovation, rewards bureaucratic inertia, and protects European firms from competition. And there are valid concerns. The precautionary principle can be vague in application and vulnerable to politicization. It can stifle early-stage technologies, delay life-saving innovations, and lead to regulatory capture by incumbents.

But the American risk-based model is no panacea. It has often privileged economic growth over public safety, whether in opioid regulation, water quality, or tech monopolies. The long-term costs of deregulation—from Flint to Facebook—are now hard to ignore.

Europe's gamble is that front-loading deliberation may slow development, but it builds public trust, international credibility, and democratic resilience. The U.S. gamble is that racing ahead will secure dominance, and that any harms can be mitigated later through torts or technical fixes. The question is no longer who is more pro-innovation—but what kind of innovation we’re willing to live with.

Precaution as Democratic Governance

One strength of the precautionary approach is that it foregrounds the public—through mandatory transparency, public comment, and multi-stakeholder review. It treats uncertainty not as a liability to be managed behind closed doors, but as a shared condition requiring social deliberation.

This was evident in Europe's response to COVID-19. While both the U.S. and EU faced failures, European countries generally adopted more coordinated and transparent vaccine policies, and maintained higher levels of public trust in their public health institutions. The precautionary mindset—combined with universal healthcare systems—meant that Europe was better positioned to prioritize the collective good without descending into misinformation-fueled gridlock.

If precaution can be misused as a cudgel, risk-based frameworks can be weaponized as excuses. Calls to “wait for more evidence” often become reasons to do nothing until it's too late. As AI systems increasingly mediate elections, military decisions, and human intimacy, delay is no longer a neutral act.

Toward a New Synthesis

The best path forward may be neither maximalism nor paralysis. It is to treat precaution not as ideology but as a governance toolkit—one that can be tailored to context, informed by democratic values, and updated through feedback. It is not anti-innovation to ask whether a technology should exist, or how it should be designed. It is anti-democratic not to.

Concrete reforms could include:

• Requiring explainability and auditability for high-risk AI systems;

• Expanding regulatory sandboxes that allow for controlled testing;

• Creating independent bodies for algorithmic review (modeled after the European Medicines Agency);

• Embedding civic participation in regulatory design, especially for marginalized communities likely to bear the brunt of harm.

Precaution is not a brake on history. It is a steering mechanism.

Then and Now

In 2001, I ended a paper with the observation that globalization seemed poised to choose one policy model—America’s or Europe’s—as dominant. I feared the incompatibility of approaches would harden into conflict. Today, that tension remains. But history has shown that absolute models rarely endure.

The real challenge is pluralism: to reconcile innovation with reflection, scale with consent, and speed with sustainability.

A conscious policy for unconscious systems is not just about AI—it’s about us.

Dennis Morgan

UST301 Summer 2001

What is most different about Europe?

A comparative look at the urban landscape of Europe and the United States from a macroscopic perspective is the easiest way to demonstrate the difference between the United States and Europe.  Policies are what shape the development of city life.

Europe has had a much longer history of urbanization.  The center generally dates to the middle age with narrow streets.  The first ring brought wider streets and a central train station in the nineteenth century.  An outer ring came with a more car friendly design, similar to America, after World War II.

It is important for this comparison to understand the place these regions hold in the broader world.  In the United States 1% of world population consumes 6.4% of total world energy.  In Western Europe, the same 1% consumes 2.4% of the world total.  Common notions of the political importance of various regions are supported by the figures of consumption (see appendix I).  Most analysis places the United States at the top of the spectrum with Europe at the next level.  Africa is at the bottom of all measures.  In general terms America and Europe are the most developed.

The disparity in energy consumption is the first macroscopic difference.  The per capita difference reflects the different way people live their lives.  These totals include computers, lights, automobiles, and flashing video recorder lights.  The numbers speak louder than words; Europeans on average use less energy than Americans.  Several factors can help explain these differences, but the effect on quality of life is difficult to quantify.  There does not seem to be a sense of deprivation in Europe, simply put long term policies have driven these realities.

The next difference is best described in the following quote, “An American traveler flying into either of the two international airports of Paris observes an unfamiliar sight: hundreds of small farms ringing the city.  On the way back to New York City, the traveler flies over a different landscape: some potato fields and an occasional boutique vineyard on the East End of Long Island’s Suffolk County, more than a hundred miles away from Manhattan.”  The value of farmland is directly influence by the level of subsidies.  The greatest obstacle to negotiations within the World Trade Organization is the conflicting historical level of support.  The data below from 1993-1995 reflects perhaps the high water mark for the US and European subsidies; negotiations are aimed at reducing the level of support to farmers.  However, these levels reflect trends over the last 100 years.

Farmland has been more valuable in Europe and has tended to slow the pace of development on the periphery of the city.  It is a mistake to think that Europe does not have Urban Sprawl.  The difference is the pace and economic incentives available to developers.  Europe, through policy decisions, has made farmland more valuable and developers have been encouraged to work within the urban core.

Low-density development is discouraged in several ways.  Bedroom communities, developed around US cities, are more difficult to heat than higher density developments.  The difference is in the economic feasibility:

Beyond the higher cost to homeowners in low-density developments, the cost of travel has been affected by the European policies that seek to reduce demand for car use.  Developments created around the automobile are less prevalent in Europe due to these policies.

European car use continues to increase, but again the major difference is the pace and scope of usage.  Car trips are increasing in Europe outside the urban core.  The Center of the cities have strong public transportation systems that seek to improve service to draw increased rider-ship.

The European governments have tended to take a more balanced approach to spending on transportation.  In the building of more car friendly areas after WWII the planners quickly recognized the difficulties faced in continued growth of car use in the urban core.  The density of structures and narrow streets made it impossible to mirror the development of the US without wholesale destruction of historical spaces.  They also faced difficulties encouraging people to use public transportation, as has been much publicized in the United States.  However, European governments have tended to focus on increasing the quality of service.  This can best be demonstrated by the more balance approach to transportation funding.

The American government has never taken such a balanced approach.  There are quality public transportation systems in the US, but one must wonder how different these systems might have been had public transportation funding been closer to parity levels seen in Europe.

It is clear that expectations of government is not the same in the United States as it is in the European context.  For example, European governments saw vacations as an essential part of any workers lifetime.  In contrast, Americans have never seen this type of issue within the realm of the government.  The disparity seen below was created through policies that were continually expanded in Europe.

At the macroscopic level, the divergence between the American and European experience could not be clearer.  These differences actually lead one to wonder what is common to these two regions.  The growing gap between the current Bush administration and the European Union seems almost more a consequence of historical trends then any particular mistake in judgment.  

The macroscopic differences seem to indicate that the course ahead for globalization is to declare one policy superior to the other.  They seem to be incompatible as represented by difficult WTO negotiations, but that is beyond my scope here.