
This site began with a question auditors are trained to ask: Who checked this, and were they independent of the people who made it?
I am a CPA. I started my career in public accounting, working on audits, and for the past two decades I have practiced in tax and wealth management. The work teaches one habit above all others. You don’t accept an answer because it is clean. You ask who prepared it, who checked it, and what happens if the check comes back “no.”
In early 2026, I turned that habit on artificial intelligence.
The first questions were not about catastrophe. They were quieter and closer to home: what happens to human thinking when a system that never hesitates begins answering every question a person used to answer alone?
That work became digitalhumanism.ai. It found that cognitive sovereignty, the ability to own your reasoning, form your own judgments, and tolerate the pause between a question and its answer, is not guaranteed. It is practiced, and practice requires friction. The systems being deployed at scale are built to remove friction. Nobody built the Smooth World on purpose. The people who built it were solving real problems. But the instruments they used were calibrated to find the bomb, not the radiation.
The longer I examined the effects, the more an auditor’s question kept returning. If these systems shape how people think, who verifies the systems? Who decides they are safe, aligned, or ready, and is that person independent of the company that built them?
The answer, again and again, was the company itself. The same institution that builds an AI system also trains it, evaluates it against its own standard, and reports the result. I call that the closed loop. In accounting, a loop like that would never pass. The profession learned, at great cost, that a company cannot audit itself.
Thinking Sovereignty was created during the 2026 tax season to examine that loop.
The site has grown well beyond its first pages. It now covers four connected areas.
AGI asks who defines artificial general intelligence, and why that definition matters to law. Alignment examines how companies decide what their systems value, and who oversees those decisions. Verification asks what it would take to independently check an AI system, and where checking ends and judgment begins. Governance measures the laws being written now, including California’s and Congress’s, against the safeguards accounting built after Enron. The Forensic Record keeps the evidence: transcripts, fabrication reports, and the documented failures behind the analysis.
Two sentences govern all of it. The AGI conversation has been conducted behind closed doors by the people who built the door. And alignment without external verification is not alignment. It is intention, and intention is not an auditable standard.
The method borrows from audit practice. Claims are traced to primary sources: statutes, filings, published research, and company documents. Every substantive page sorts its findings into three tiers: what the evidence demonstrates, what can reasonably be inferred, and what remains open. When a page goes beyond the record, it says so.
AI systems are part of the method and part of the subject. ChatGPT, Claude, and Gemini have been examined under sustained questioning, and the most important transcripts are published in full. The same systems also helped build these pages. That makes their contributions a separate check, not an independent one, and the pages say so plainly.
When a page is wrong, it is corrected in place, with a dated editor’s note.
The goal is simple to state and hard to achieve: a public record that holds AI to the standard every audited company already meets. Not trust, but verification. Not a company’s description of its safeguards, but evidence that someone independent tested them.
The site also offers alternatives, not only criticism. Its model statute, the Independent AI Provenance and Examination Act, is one attempt to show what independent examination could look like. It is a proposal, tested on this site by the same questions applied to every law.
Jim Germer is a CPA with more than three decades in the profession, the founder of The Human Choice Company LLC, and the author of thinkingsovereignty.ai and digitalhumanism.ai. His writing on AI and audit independence appears under his own byline in Accounting Today, CPA Practice Advisor, and TEACH Magazine. He does not currently perform or solicit AI audits. If that changes, I will disclose it here.
Stay Sovereign.
Jim Germer
© 2026 Jim Germer - The Human Choice Company LLC. All Rights Reserved.
We use cookies to improve your experience and understand how visitors use our website so we can make it better.