‘The Intelligence Republic: AI Independence and the Future of Self-Government'

Book Notes #238

By Andrew Roth, Scholar-in-Residence

Have you ever heard of “Murphy’s Law(s)?”

There are many “Murphy’s Laws” or, one might say, aphorisms masquerading as a “Murphy’s Law.” Attributed to aerospace engineer Edward Murphy, the original Murphy’s Law states that “Anything that can go wrong will go wrong.” [1]

Which, when considering artificial intelligence’s potential societal impact, is a maxim worth remembering.

Another alleged “Murphy’s Law” asserts that “Someone always knows.” I have no idea who first said it. Still, any speaker ought to be aware that in any audience there will almost always be someone who knows more about the topic than they do. It is a thought I keep front of mind whenever I speak to any group. Once, at The Chautauqua Institution, after giving a presentation on “Sports and the Quest for Women’s Rights,” a woman came up to me and introduced herself as Patricia Ireland, former president of the National Organization for Women (NOW).

In the summer of 2025, presenting a series of lectures for Chautauqua’s Roads Scholars Program on the future of the American Experiment, I met Jon Neiditz. It turns out he and I share a passionate interest in the future of the American republic and how its currently polarized politics threaten its future. Although a third of my presentations focused on storytelling and media and technology’s impact on American politics and society, I barely scratched the surface about AI’s potential to either refine or shred that story’s many threads.

Which, it turns out, was a good thing, because Neiditz, proving “Murphy” prophetic, is an expert on the topic. A retired partner at the Atlanta law firm Kilpatrick Townsend, Neiditz founded the firm’s Cybersecurity, Privacy and Data Governance Practice. As stated on the firm’s website, he “remains the only person recognized by The Best Lawyers in America® both for Information Management Law and for Privacy and Data Security Law, also including him for Technology law and as the 2023 “Atlanta Lawyer of the Year” in Electronic Discovery and Information Management Law. [2]

Based upon his passion for the future of the American republic and the rapidly evolving impact of artificial intelligence upon each, Neiditz just published a brilliant short book outlining AI’s challenge and how society can meet it – The Intelligence Republic: AI Independence and the Future of Self-Government.

While much of the discussion surrounding the rapid deployment of AI focuses on either the immediate and practical (“Will it eliminate my or my grandchildren’s jobs?”) or the ethereal (“Is AI a conscious entity and will it conspire with other AIs to take over the world?”), Neiditz is interested in its impact upon individuals’ well-being and ability to function as citizens.

People’s ability to exercise their duties and enjoy their rights as citizens is the essence of republican government. As we have discussed repeatedly in these Book Notes, the foundational principle of the American republic is government legitimized by the consent of the governed. As AI becomes ever more deeply embedded in all facets of society, the central political question of our time is not whether AI will be used but whether the systems that AI powers will remain answerable to the people they impact.

Some mundane but not trivial examples Neiditz touches on illustrate this point.

Have you ever had your doctor’s prescribed course of treatment overridden by an insurance company’s algorithm, requiring you to try a different treatment before receiving the one your doctor recommended? Who knows more about you: the insurance company or your doctor?

Have you ever sought advice about how to use a new product or tried to lodge a complaint about the service you received only to be caught in an online “Chat” doom loop? Then, when you requested to speak to an actual human, you were rerouted to a phone conversation with another chatbot?

Or, if you are a teacher, have you ever received a written assignment from a student clearly beyond the student’s ability, even if the submission itself was written in noun clusters and a numbingly passive voice?

Or, more ominously, have you ever had your work performance monitored by a system tailored to measure tangible metrics but utterly oblivious to the intangible nuances of interpersonal interactions affecting output?

Or, a political question, in an AI saturated world where not only language and text can be manipulated but also images and videos, have you ever asked yourself about the truth or falsity of the information you need to make the informed citizen decisions democracy demands?

I suspect you could supply your own specific example for each, because each is a real-world example of AI reducing you to an item to be measured rather than a person to be served. In short, you’ve lost (or, more ominously, maybe ceded) decision making about your own life to an algorithm.

The larger social implication, as Neiditz says, is that “A people that cannot govern its intelligence cannot govern itself.” [3]

So, how are a people to gain control over the information systems that increasingly shape their world?

The first thing is to understand the problem. From Neiditz’s perspective, the problem has four dimensions: 1) life by permission; 2) AI as mediation; 3) “The Gentle Cage;” and 4) “unanswerable necessity.”

In “life by permission,” both organizations and individuals depend upon systems over which they have no control. An organization might lease an Enterprise Resource Planning System (ERP) to manage their operations but not have access to the system’s codes. Therefore, they must fit their processes to the ERP’s structure. Should they alter the codes, the vendor can “flip the switch” turning the system off. Similarly, on an individual level, you can only operate as the software you employ permits. You can’t change its procedures, and if you try to move to another platform, you will likely discover that your data is not transferable. Neiditz describes this as living “by permission.” [4]

That is perilously close to the rentier society Peter Frase described in Four Futures, which I discussed in Book Notes #217 and can be found here. A rentier society is essentially the landlord/tenant model writ large. You do not own or control anything. You have access, you have permission to use a tool, a home, whatever, but only according to the owner’s rules and protocols. Most importantly, that permission, that access, can be withdrawn. According to Neiditz, access is not freedom, and freedom is the foundation of republican government.

From Neiditz’s perspective, AI is the ultimate “middleman,” to use an old term. If the internet itself was the ultimate harbinger of disintermediation — the economic term describing a system of direct exchange between purveyor and consumer — then AI is the “middleman’s” return with a vengeance. AI does not need to issue direct commands to exercise power over your choices. It mediates between you and your choices. And the ability to choose is by definition the exercise of freedom.

So, when you allow AI to rank your options according to its own criteria rather than your own, you begin to cede your power to choose. In doing so, AI – Copilot, ChatGPT, Claude, or the cute little “helper” icon on whatever website you’re visiting – begins to shape the experience you call “life.” Neiditz calls this power “the architecture of possibility.” [5]

As a result, echoing Aldous Huxley’s Brave New World and Neil Postman’s Amusing Ourselves to Death, Neiditz observes that AI’s greatest danger does not come via force. It comes as convenience, even entertainment. He calls this “the gentle cage” as AI slowly relieves us of the burden of thinking for ourselves. [6] As it becomes habitual, it quietly weakens our ability to understand or challenge important decisions. The system does not command. It does not tell you that you can’t do this or that but slowly shows you that you need not as it summarizes a document, writes (edits) an email, and makes a choice for you eliminating the burden of judging for yourself. The danger is not in the assistance AI provides, but in its potential to gradually erode your ability to decide for yourself. In so doing, you cede a bit of your freedom click-by-click.

The problem is not AI itself, but that it becomes “wedged into necessary life” as an “unanswerable intelligence” guiding our decision-making processes. [7] It slowly becomes something people cannot do without but also can neither challenge nor easily replace. He says such a situation arises when a system becomes both indispensable and unanswerable. On an individual level, as already noted, it raises the question “Are you making your own choices?” On the organizational level, once an organization cannot realistically function without an AI system, that AI system ceases to be an ordinary commercial product. It begins to resemble a form of sovereignty or power raising the questions “Who is in charge here?” “Who is calling the shots?”

As a corrective, Neiditz posits what he calls an Intelligence Republic. In such a society, rejecting calls for the government to either own it (as he says, “a share is not a say”) or to declare it a public utility, he thinks AI can remain privately owned. But in such a society, in order to retain meaningful authority over the technology shaping their lives, citizens must have four powers or rights.

First, they must have a right to “see” the machine, i.e., know that a decision affecting their life was made by a machine. [8] The machine’s authority to decide should not be disguised as simple procedure or operational necessity. Nor should its decision-making ability be hidden behind proprietary claims protecting its code, although its code need not be revealed. All that’s required is enough of the protocol be disclosed so that the person impacted can understand what happened and what information guided the decision.

Secondly, the affected person should have the right to question the machine/algorithm and to get an answer from “a person who will stand behind” it. They should have the right to appeal to a human “who can do more than admire” the machine’s answer but actually intervene to either fully explain the decision and/or revise it based upon new information or corrected information provided by the person impacted by the machine’s decision. [9] In the example given earlier about an insurance algorithm overriding a doctor’s recommendation, the affected patient should have the right to challenge that decision not to another machine-driven algorithm but to an actual person empowered to make decisions, explain them, and be held accountable for them.

Next, if freedom is the ability to choose, then one of the options must be the right to refuse, which Neiditz calls the power to leave. [10] Using the medical example above, the power to leave must include the right to seek an alternative without forfeiting your medical records; if it is an educational issue, the power to transfer and take your transcript with you; if an employment issue, the power to leave and seek similar employment. In all instances, it must include the ability to seek alternatives.

Lastly, Neiditz includes a fourth power or right. It, however, comes at the beginning. In all instances, but particularly in civic instances, people must have not only the power to react but also the power to decide before a system is deployed. People must have the right to be involved in the determination of the AI system’s purpose. These questions are not technical; they are political. Since they are not technical, they are not the purview of engineers. Since they are political, and what is politics but the process of choosing what is appropriate for an individual, an organization, or a society, these questions must be answered by the people who the AI system will impact. [11]

How is this to be done?

At both the organizational and governmental level, it is the age-old question of who decides. Neiditz offers an institutional remedy based on a governmental model that could also, in many instances, be applied to organizations. But it is essentially a governmental model. He rejects turning over the government to the corporations who own the technology and the opposite total technological control by the state. He notes that a weak government might rent its operating systems from private companies that it cannot effectively supervise. Conversely, an authoritarian government may use AI to control its citizens.

In a chapter titled “Power That Answers, From Town to World,” [12] Neiditz describes his vision of a governmental authority strong enough to establish rights and create alternatives when systems fail. But, and it’s an important “but,” it is a governmental authority that must remain divided. Based on the American concept of federalism, AI governance should take place at the lowest level that is both capable and accountable. Schools, libraries, hospitals, towns, and counties should develop enough technical knowledge to evaluate and sometimes reject AI systems. Local institutions may join consortia to share expertise and bargaining power. The national government should provide common rights and infrastructure security without becoming the sole owner or operator of American intelligence.

To that end, revealing his hopes for the continued American Experiment and modeled on the Declaration of Independence, whose 250th anniversary we just celebrated, Neiditz says government and societal AI independence does not mean independence from AI. It means independence from an AI that cannot be challenged. To ensure that AI remains a tool of citizens and citizens do not become simple tenants of a machine guided society, Neiditz’s Intelligence Republic declares its independence from:

“No private sovereign over public intelligence.
No public power that cannot say why.
No verdict without a person who will answer for it.
No dependence without a door.
No machine that deceives us about what it knows, what it does not know, or whose side it is on.
No comfort sold to us in exchange for our judgment.
No share offered in the place of a say.
No sky enclosed above a free people and rented back to them.
No necessary thing permitted to become a master.
No self-government by permission.” [13]

I have just skimmed the surface of Neiditz’s attempt to provide a map for how both AI might be harnessed for the common good, and in the process, the American Experiment protected.

It is a thought-provoking read. One I encourage anyone wrestling with the implications of AI for the future of American culture to read. It does not provide easy answers. But it does provide a cogent and clear analysis of the dangers AI presents and a simple approach for taming them. Neiditz doesn’t say it will be easy. The work of freedom, the work of democracy never is, and it begins with you.

---

The Intelligence Republic: AI Independence and the Future of Self-Government” can be purchased at here.

To further sharpen your leadership skills to meet the challenges of developing sound AI policies, consult my book recently published by the Jefferson Educational Society “How to Be a Highly Effective Leader: A Primer.” It can be purchased at www.jeserie.org/books or by calling 814-459-8000.

---

Photo Credits

“Cover of ‘The Intelligence Republic: AI Independence and the Future of Self-Government” at Amazon.com available at The Intelligence Republic: AI Independence and the Future of Self-Government - Kindle edition by Neiditz, Jon. Politics & Social Sciences Kindle eBooks @ Amazon.com. accessed July 13, 2026.

End Notes

1. “Murphy’s Law,” at Encyclopedia Britannica available at Murphy’s law | Adage, Origin Story, Variations, & Corollaries | Britannica accessed July 13, 2026.

2. “Jon Neiditz,” at Kilpatrick available at Jon Neiditz accessed July 13, 2026.

3. Neiditz, Jon. The Intelligence Republic: AI Independence and the Future of Self-Government (Atlanta, GA: Jon Neiditz, 2026), p. 6. All citations are from a PDF copy provided by the author and may not correspond to the published version.

4. Ibid., p. 6.

5. Ibid., p. 10.

6. Ibid., p. 13.

7. Ibid., p. 19.

8. Ibid., p.25.

9. Ibid., p.26.

10. Ibid.

11. Ibid., p. 38 ff.

12. Ibid., pp. 44-47.

13. Ibid., pp. 52-53.

---

Andrew Roth, Ph.D., is a Scholar-in-Residence at The Jefferson Educational Society. Reach him at roth@jeserie.org.

---