When Neutrality Is Not Neutral: Lessons for Legal Risk Assessment in Polarised Environments

The Illusion of the View From Nowhere

The managing director of Verfassungsblog, one of Europe's most respected public law platforms, spent last weekend physically blocking delegates of the AfD from reaching their party convention in Erfurt. He has written openly about whether that was appropriate for someone in his institutional role. His answer, in essence, is that false equivalence between constitution-loyal and constitution-hostile actors is itself a political choice, even if it presents itself as principled neutrality.

That argument deserves careful attention far beyond the world of legal academia. It surfaces a tension that in-house counsel and law firms navigate constantly, often without naming it: the assumption that legal analysis is valueless, that spotting risk is a mechanical exercise, and that the framing of advice is separable from its substance.

It is not.

Framing Is a Legal Skill, Not a Political Luxury

When a contracts team reviews a counterparty operating in a jurisdiction where democratic norms are visibly eroding, the question is rarely whether the contract is formally valid. The question is what category of risk the relationship represents over the life of the agreement. Force majeure clauses, governing law provisions, dispute resolution mechanics, and termination triggers all look different depending on whether the drafter treats institutional stability as a fixed background assumption or as a variable to be stress-tested.

Neutrality in that context does not mean treating all jurisdictions identically. It means applying consistent analytical rigour, which will produce different outputs in different environments. A clause that performs reliably under English or German law may become unenforceable or practically worthless if the court system in the relevant jurisdiction has been subordinated to executive discretion. Treating that difference as a political opinion rather than a legal fact is the kind of false neutrality the Verfassungsblog piece is really arguing against.

In-house teams that flag this category of risk clearly serve their organisations better than those who leave the inference to the business. That is not advocacy. That is legal craft.

What AI Contract Tools Must Understand About Context

This is precisely the domain where generic AI contract tools fall short. A system trained to identify standard clause structures, compare against market precedents, and flag deviations from a playbook will not, by default, ask whether the governing law of this agreement is law in a meaningful sense in three years' time. It will not notice that an arbitration seat chosen for its perceived neutrality has recently passed legislation curtailing judicial independence. It will not read the contract from your side of the table with awareness of where geopolitical and legal risk is accumulating.

Adira is built on a different premise. Reading contracts from the client's perspective, in the client's voice, and with knowledge of the law as it actually operates in the relevant jurisdiction means the system must carry a model of legal context, not just legal text. Jurisdiction is not a metadata field. It is a substantive input that shapes what every clause in an agreement actually means and what it will be worth if enforced.

The Erfurt story is a reminder that legal scholars themselves disagree about where analysis ends and judgment begins. AI tools should be honest about that boundary rather than hiding it behind an appearance of mechanical objectivity.

Practical Implications for Contract Governance

For in-house teams and external counsel, the operational takeaway is straightforward. Contract governance frameworks should include an explicit political and institutional risk layer alongside the more familiar commercial and legal risk categories. That layer should ask: is the legal infrastructure underpinning this agreement stable, and if not, what contractual protections can substitute for institutional reliability?

This means revisiting boilerplate with fresh eyes. Stabilisation clauses, which fix the applicable law at the date of signing, matter more in volatile jurisdictions. Hardship provisions that reference systemic changes to the legal environment deserve more attention than they typically receive. Choice of seat in arbitration should reflect a genuine assessment of institutional independence, updated periodically rather than inherited from the last deal.

None of this is exotic. It is standard practice in emerging market transactions. The unsettling development of the past decade is that the analysis now needs to reach further into markets that were once considered stable.

Conclusion: Judgment Cannot Be Outsourced, But It Can Be Structured

The Erfurt debate will continue in German public law circles for some time. For the legal technology sector, the more immediate lesson is that tools built to eliminate judgment from contract review are solving the wrong problem. The goal is to structure judgment: to ensure that the right questions are asked consistently, that context is not stripped away in the pursuit of scalability, and that the advice produced is genuinely read from the client's side of the table.

Neutrality, applied thoughtlessly, is its own form of analytical failure. Good contract AI, like good counsel, knows the difference.