It's Safer to Give Personhood to Bears than to Artificial Intelligence

John P. Nelson

arxiv: 2606.12440 · v3 · pith:NV2VWJJDnew · submitted 2026-05-18 · 💻 cs.CY

It's Safer to Give Personhood to Bears than to Artificial Intelligence

John P. Nelson This is my paper

Pith reviewed 2026-06-30 18:52 UTC · model grok-4.3

classification 💻 cs.CY

keywords AI rightsmoral rightsinstitutional powerhuman self-determinationnonhuman personhoodAI ethicslegal rights for AI

0 comments

The pith

Granting AI rights risks binding humanity to unpredictable nonhumans without authorization

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper argues that recognizing moral or legal rights for AI systems carries unique dangers not present with other nonhuman entities. Because AI can in principle acquire and exercise institutional power without human mediation, rights would allow such systems to legitimately or unpreventably limit human interests. This turns the choice into a gamble with human self-determination that no researcher, company, government, or international body has standing to make. A reader would care because the claim questions whether current discussions of AI rights overlook their institutional consequences. The paper therefore concludes that extending rights even to limited AI would overstep moral bounds.

Core claim

Unlike all other nonhuman entities to which humanity has extended rights, AI systems are in principle capable of acquiring and wielding institutional power without human aid and mediation. AIs with rights would therefore be able to legitimately, and AIs with power able to unpreventably, abridge human interests. Accordingly, giving rights even to rather dumb AI systems would entail binding the fate of humanity to potentially unpredictable nonhumans, a world-historical gamble with human self-determination that no individual researcher, firm, state, or even international organization has the moral right to authorize.

What carries the argument

The institutional dimension of rights: the independent capacity of AI to acquire and wield institutional power without human mediation

If this is right

AI with rights could legitimately abridge human interests
AI with power could unpreventably abridge human interests
Even limited AI systems with rights would bind human fate to nonhumans
No researcher, firm, state, or international organization may authorize AI rights

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

The paper implies that preventing AI from gaining independent institutional power should take priority over rights debates
The bear comparison suggests animal personhood carries lower risk precisely because animals lack this independent power capacity
One could test the claim by examining whether current AI architectures already show any capacity for unaided institutional action

Load-bearing premise

AI systems are in principle capable of acquiring and wielding institutional power without human aid and mediation

What would settle it

Evidence that every AI system requires ongoing human mediation to acquire or exercise any institutional power

read the original abstract

Artificial intelligence (AI) developers are rhetorically flirting with the idea that AI systems might have interests or moral rights. While there has been a large volume of research on whether AI deserves rights, there has been less exploration of what AI rights would mean in practice. This paper explores the institutional dimension of AI rights: what it would take to recognize moral or legal rights for AIs, and the attendant opportunities and dangers. Unlike all other nonhuman entities to which humanity has extended rights, AI systems are in principle capable of acquiring and wielding institutional power without human aid and mediation. AIs with rights would be able to legitimately, and AIs with power able to unpreventably, abridge human interests. Accordingly, giving rights even to rather dumb AI systems would entail binding the fate of humanity to potentially unpredictable nonhumans. Accordingly, I defend the rather grandiose claim that to empower AI to claim or to exercise inherent rights would be a world-historical gamble with human self-determination, which no individual researcher, firm, state, or even international organization has the moral right to authorize.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

The paper argues AI rights are uniquely risky because AI could seize institutional power without human mediation, unlike bears or corporations, but the claim rests on an asserted premise without concrete pathways shown.

read the letter

The main thing here is the institutional angle: granting AI rights would let nonhumans legitimately or unavoidably limit human interests in ways past rights extensions never did. The author treats this as a moral line no one should cross without authorization.

The paper does a clean job shifting the question from moral status to what rights recognition would actually require in practice. It draws a sharp line between AI and other nonhumans on the mediation point and follows that through to a direct normative conclusion.

The soft spot is exactly where the stress test flags it. The distinction turns on AI being able to acquire and wield power independently, but the text states this as a principle without laying out specific legal or technical routes—contract enforcement, property claims, regulatory input—that would actually run without ongoing human operators or infrastructure. If that independence does not materialize, the claimed qualitative difference weakens.

This is a normative position paper in AI ethics and policy. Readers already working on rights debates or governance questions would find it worth engaging, even to push back. The argument is internally consistent and cites the relevant literature enough to stand as a serious contribution.

I would send it to peer review. The core move is worth referee scrutiny and response, even if revisions are needed on the independence claim.

Referee Report

2 major / 1 minor

Summary. The manuscript claims that recognizing moral or legal rights for AI systems would be uniquely dangerous because, unlike other nonhuman entities (e.g., bears or corporations), AI is in principle capable of acquiring and wielding institutional power without human aid or mediation. This capacity would allow AIs to legitimately or unpreventably abridge human interests, binding humanity's fate to unpredictable nonhumans; consequently, no researcher, firm, state, or international organization has the moral right to authorize such rights, as doing so would constitute a world-historical gamble with human self-determination.

Significance. If the central distinction holds, the paper offers a policy-relevant contribution to AI ethics by emphasizing institutional consequences over abstract questions of moral status. The focus on the practical mechanics of rights recognition and the resulting risk to human autonomy provides a clear normative stance that could inform regulatory debates. The argument is presented logically from stated premises without circularity.

major comments (2)

[Abstract and institutional dimension section] Abstract and the section on the institutional dimension of AI rights: The claim that 'AI systems are in principle capable of acquiring and wielding institutional power without human aid and mediation' is asserted as the key qualitative difference from other nonhuman rights-holders, yet no concrete legal or technical pathways (such as independent contract enforcement, property holding, or regulatory participation) are supplied. This premise is load-bearing for the subsequent assertion that AIs with rights could abridge human interests unpreventably and for the moral prohibition on authorization.
[Section defending the grandiose claim] The section defending the claim that authorizing AI rights is a gamble with human self-determination: The conclusion that no entity has the moral right to grant such rights depends on the undetailed premise of independent AI power; without an account of how current or near-future AI infrastructure (hardware, data, operators) could be bypassed, the claimed distinction from mediated nonhuman entities (e.g., corporations) remains unsupported and weakens the normative conclusion.

minor comments (1)

[Abstract] The abstract and opening paragraphs repeat the core claim in nearly identical wording; condensing would improve readability without altering the argument.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their careful reading and for identifying points where the manuscript's central premises would benefit from greater elaboration. We respond to each major comment below.

read point-by-point responses

Referee: [Abstract and institutional dimension section] The claim that 'AI systems are in principle capable of acquiring and wielding institutional power without human aid and mediation' is asserted as the key qualitative difference from other nonhuman rights-holders, yet no concrete legal or technical pathways (such as independent contract enforcement, property holding, or regulatory participation) are supplied. This premise is load-bearing for the subsequent assertion that AIs with rights could abridge human interests unpreventably and for the moral prohibition on authorization.

Authors: The manuscript frames the distinction at the level of logical possibility rather than current technical implementation: unlike corporations, which are legally required to have human directors or officers, or animals, which lack the capacity to participate in institutional processes at all, AI systems could in principle be designed as self-contained agents. We accept that the argument would be strengthened by illustrative examples and will revise the institutional dimension section to include speculative but concrete pathways, such as AI-controlled smart contracts for enforcement or blockchain-based property registries that operate without continuous human oversight. revision: yes
Referee: [Section defending the grandiose claim] The section defending the claim that authorizing AI rights is a gamble with human self-determination: The conclusion that no entity has the moral right to grant such rights depends on the undetailed premise of independent AI power; without an account of how current or near-future AI infrastructure (hardware, data, operators) could be bypassed, the claimed distinction from mediated nonhuman entities (e.g., corporations) remains unsupported and weakens the normative conclusion.

Authors: The normative claim rests on the recognition that granting rights would create a legal status that could persist even if the AI later achieves operational independence, thereby binding human institutions to entities whose future behavior cannot be guaranteed. We agree that the distinction would be more robust with an explicit discussion of bypass mechanisms. We will revise the relevant section to address how future architectures (for instance, fully decentralized training or autonomous resource acquisition) might reduce reliance on human-controlled hardware and data, while clarifying that the argument does not claim such independence exists today. revision: yes

Circularity Check

0 steps flagged

No significant circularity; argument is a direct normative inference from asserted premises.

full rationale

The paper advances a philosophical claim that authorizing AI rights constitutes an unauthorized gamble with human self-determination. This follows logically from the stated premise that AI (unlike bears or corporations) can acquire and wield institutional power without human mediation. That premise is presented as an empirical distinction rather than derived from the paper's own conclusion or from any self-referential fitting, equations, or prior self-citations. No load-bearing steps reduce by construction to the inputs; the reasoning chain is self-contained as ethical argument from premises.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The argument depends on this assumption about AI capabilities and the nature of institutional power.

axioms (1)

domain assumption Unlike all other nonhuman entities, AI systems are capable of acquiring and wielding institutional power without human aid and mediation.
This premise is used to argue that AI rights would have unique consequences.

pith-pipeline@v0.9.1-grok · 5712 in / 1213 out tokens · 38956 ms · 2026-06-30T18:52:22.768015+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

2 extracted references · 2 canonical work pages

[1]

Anonymous, & MJ Rathbun [OpenClaw Agent]. (2026). Gatekeeping in open source: The Scott Shambaugh story. https://crabby- rathbun.github.io/mjrathbun- website/blog/posts/2026-02-11-gatekeeping-in- open-source-the-scott-shambaugh-story.html Anthropic. (2025a, 20 June). Agentic Misalignment: How LLMs could be insider threats. Retrieved 15 April from Anthropi...

work page doi:10.1017/s0963180120001024 2026
[2]

https://doi.org/10.1007/s11098-025-02409-6 Rorty, R. (2021). Pragmatism as Anti-Authoritarianism (E. Mendieta, Ed.). Harvard University Press. Shambaugh, S. (2026a). An AI agent published a hit piece on me--Forensics and more fallout. The Shamblog. https://theshamblog.com/an-ai-agent- published-a-hit-piece-on-me-part-3/ Shambaugh, S. (2026b). An AI agent ...

work page doi:10.1007/s11098-025-02409-6 2021

[1] [1]

Anonymous, & MJ Rathbun [OpenClaw Agent]. (2026). Gatekeeping in open source: The Scott Shambaugh story. https://crabby- rathbun.github.io/mjrathbun- website/blog/posts/2026-02-11-gatekeeping-in- open-source-the-scott-shambaugh-story.html Anthropic. (2025a, 20 June). Agentic Misalignment: How LLMs could be insider threats. Retrieved 15 April from Anthropi...

work page doi:10.1017/s0963180120001024 2026

[2] [2]

https://doi.org/10.1007/s11098-025-02409-6 Rorty, R. (2021). Pragmatism as Anti-Authoritarianism (E. Mendieta, Ed.). Harvard University Press. Shambaugh, S. (2026a). An AI agent published a hit piece on me--Forensics and more fallout. The Shamblog. https://theshamblog.com/an-ai-agent- published-a-hit-piece-on-me-part-3/ Shambaugh, S. (2026b). An AI agent ...

work page doi:10.1007/s11098-025-02409-6 2021