What We Oweto Each Other

The First Collision

In February 2026, while sitting in my office, I opened a Flipboard notification about a confrontation between Anthropic and the Pentagon in an AP article.

The dispute concerned the conditions under which the Department of Defense could use Claude. Anthropic refused to remove safeguards related to mass domestic surveillance and fully autonomous weapons, arguing that current frontier systems remained too unreliable for certain high-stakes uses. Pentagon officials responded that a private company should not determine how the military could lawfully use technology it had purchased. Both sides claimed to be protecting national security. Both institutions possessed something the other could not easily replace: the government held public authority, while Anthropic held technical expertise and control over the system.

At first, I read it as a dispute over one government contract, but it exposed a much larger institutional problem. AI is being built in one world and governed in another. The people developing the systems, the people legally empowered to make public decisions, and the people who will bear the consequences of failure are often not the same people.

As a student, I had already watched ChatGPT and Claude develop at a pace that was difficult to reconcile with the slower world of laws, public institutions, and university disciplines. Their rapid improvement made me wonder whether an equivalent policy world existed around them. But the better question was not whether AI policy existed. It clearly did. The real questions were more difficult:

1. Who has legal or political authority?
2. Who controls the technology or infrastructure?
3. Why is there no reliable institution connecting these groups?

The Anthropic-Pentagon dispute gave institutional form to a problem I had previously understood only in the abstract. Governments cannot govern advanced AI without access to technical knowledge. AI companies cannot independently determine the public rules governing defense, security, surveillance, or biological risk. And neither side can simply replace the other. Yet, the institutions connecting them remain fragmented, temporary, and frequently adversarial.

What is expected of us, then, is neither perfect coordination nor a single institution capable of seeing everything. What seems more realistic, and more useful, is to begin by seeing the relationships clearly. Who understands the technology? Who can make the decisions? I have found it helpful to think about these questions as three maps of the same world. Each reveals something the others cannot, and together they begin to show where AI and policy meet, and where they still come apart.

This is where the thought process behind Omoikane began for me. The main claim behind the project is that the central governance problem of frontier AI is the separation of technical knowledge and public accountability. The greater the distance between who knows, who decides, and who bears the risk, the more likely AI policy is to become late, adversarial, or technically unworkable. Omoikane is proposed as a policy-intelligence platform for mapping that distance across public and private institutions.

What Is Expected of Us1In Ted Chiang's 2005 short story “What's Expected of Us” (PDF), the story considers what happens when people come to believe that the future is already fixed and that their choices no longer matter. I am using it here less as an argument about free will than as a warning against the passive language that often surrounds technology. Read More in Notes

AI is often described as something that is simply happening to us, passively. Models will become more capable, competition will intensify, and governments will eventually have to adapt. This language captures the speed of change, but it can also make responsibility disappear. Before asking whether policy can keep up, we should ask a more basic question: what is expected of us? Here, I would put the problem more simply: rooms full of experts, but too few hallways between them.

Much of frontier AI is developed inside a small number of private companies. These companies recruit the technical talent, operate the computing infrastructure, evaluate their models, and control how those models are released. Stanford's 2025 AI Index found that industry produced roughly 90 per cent of the notable AI models identified in 2024. At the same time, the U.S. Government Accountability Office has warned of serious shortages of AI expertise across the federal workforce. The people closest to the technology are therefore often separated from the institutions expected to govern its public consequences.

Authority is divided in a different way. Governments distribute responsibility across legislatures, courts, defence departments, regulators, and international bodies. Each sees the technology through its own mandate. One institution may write the rules, another may purchase the system, and another may only become involved after something has gone wrong. Responsibility still exists, but it is spread across institutions that may not share the same information or even agree on what the main problem is.

The divide is also linguistic. Engineers tend to ask whether a system works and how reliably it performs. Policymakers are more likely to ask who is authorized to use it and who will be held responsible if it fails. Companies must also consider deployment, competition, and their obligations to customers or investors. They may all be discussing the same model while approaching it as a different kind of problem.

Their timelines rarely match either, as models can change within weeks, while legislation, judicial review, and international effort can take years. It is tempting to reduce this to the claim that policy moves too slowly, but caution is part of what public institutions are for. They are expected to follow procedures and remain accountable to law. Private firms can usually move faster because they are not bound by the same public processes, but speed does alone does not give them public legitimacy.

This mismatch becomes more serious when private technical choices begin to shape public life. A decision about model access or safeguards may be made inside a company, yet its consequences can reach national security, civil liberties, and critical infrastructure. Governments may then depend on systems they cannot fully inspect, meanwhile the company understands the technology but does not possess the authority to decide every public use. In return, the government possesses that authority but may lack the knowledge needed to exercise it well.

I therefore do not think the central problem is a simple shortage of expertise. The expertise exists already, but it is distributed across institutions that organize knowledge and authority separately. Specialization still matters too, because it gives us depth and makes complex work possible. The problem begins when those divisions also become the main way we understand issues that are already moving across them.

The Three Maps

What We Owe to Each Other

Then, as people working in policy and technology, we have to ask ourselves: what is expected of us? More importantly, what do we owe one another2I am using this phrase from T. M. Scanlon's What We Owe to Each Other in a institutional sense. When technical and policy institutions jointly shape systems that affect public life, neither can treat the knowledge or responsibilities of the other as someone else's problem., and what should that require of us? I do not think the answer is that engineers should become policymakers, or that policymakers need to understand every technical detail. These fields exist separately for good reasons. But once the same systems begin to shape public life, that separation can no longer excuse the distance it has.

When different institutions each hold part of the knowledge needed to understand a serious risk, they have some responsibility to make those parts intelligible to one another. The failure in this challenge is simply allowing the gaps and distance between them to remain even after their consequences have become visible.

Alignment, then, is not a matter of policy catching up with technology, or technology simply submitting to policy. It should create enough shared understanding where each side can see what the other knows, what it cannot know, and where responsibility cannot simply be passed along. The goal is not to arrive at some perfect agreement as we try to close this gap.3Scanlon's contractualism centres on whether conduct can be justified to others while respecting their separate interests, rather than securing agreement or producing a preferred overall outcome. Read More in Notes What matters is that the institutions involved can explain their decisions to one another and to the people who will live with their consequences.

Omoikane is my attempt to make those relationships easier to see. It would not tell institutions what to decide, nor would it remove disagreement between them. It would show where knowledge sits, where authority lies, and where one institution has become dependent on another. I think that matters a lot, because good judgment depends, at least in part, on having an honest picture of the system in which decisions are being made.

Perhaps this is what the worlds of policy and technology owe one another: not to become the same, but to make a serious effort to render their knowledge understandable across the boundary between them.