AI Rights & Personhood
The Next Frontier in Law, Ethics & Philosophy
JJ Shay | Global Gauntlet AI bit.ly/jjshay
THE QUESTION WE CAN'T IGNORE 66%
"Most likely, after an LLM patiently explained why it believed it was sentient, we would simply go on using it as a tool."
of consciousness researchers believe artificial consciousness is plausible under certain computational models
— Jerry Kaplan, Computer Scientist
Source: Vox 2024 Survey of AI Researchers
What Is Legal Personhood?
Legal personhood is a status that grants an entity rights and obligations under the law—the ability to sue, own property, enter contracts, and be held accountable.
NATURAL PERSONS
Human beings with inherent rights from birth—life, liberty, property
CORPORATE PERSONS
Corporations, trusts, and organizations—legal fictions created for economic efficiency
AI PERSONS?
The emerging question: Should AI have limited legal recognition in high-stakes domains?
Personhood Has Always Evolved
Roman Era
Persona ficta—legal fiction for non-human entities
1886
Corporations granted constitutional rights (Santa Clara v. Southern Pacific)
2008
Ecuador grants rights to nature in its constitution
2017
New Zealand's Whanganui River recognized as legal person
2022-24
Idaho & Utah explicitly deny AI personhood
Key Insight: Legal personhood has been extended for instrumental governance needs—not just moral agency.
The AI Capabilities Explosion
percentile on bar exam
AI welfare researcher hired (Anthropic)
Emerging AI Capabilities Raising Questions:
• Self-modeling and situational awareness
• Long-term planning and goal revision
• Epistemic memory and learning from experience
• Reasoning about abstract concepts including ethics
Many experts predict general AI
Three Frameworks for AI Personhood
RIGHTS-BASED
FUNCTIONALIST
AGENCY-BASED
Focuses on intrinsic moral worth. If AI achieves sentience or consciousness, it may deserve rights regardless of utility.
Personhood serves practical purposes. Limited AI status could solve liability gaps and enable contracts without full rights.
Emphasizes autonomous decisionmaking. AI that sets goals and acts independently may warrant moral consideration.
"No single approach fully captures AI's complex profile as both a powerful tool and a non-sentient actor."
— Technology and Regulation Journal, 2025
THE EMERGING SOLUTION
The Hybrid Model
Limited, context-specific legal recognition for AI in high-stakes domains while preserving ultimate human accountability
Where This Could Apply:
Financial Services Medical Diagnostics Autonomous Vehicles Legal Document Review
THE BIG QUESTION
Can AI ever be conscious—and would we even know?
The Two Camps:
BELIEVERS
If AI replicates the functional architecture of consciousness, it will be conscious—regardless of silicon vs. neurons.
SKEPTICS
Consciousness requires biological processes in an embodied organic subject—silicon can only simulate, not experience.
Dr. Tom McClelland (Cambridge): "The only justifiable stance is agnosticism—we simply won't be able to tell."
Consciousness ≠ Moral Status
The critical distinction: It's not just about awareness—it's about the capacity to suffer.
CONSCIOUSNESS
Perception, self-awareness, ability to model the world—can still be ethically neutral. Self-driving cars that experience the road wouldn't require moral consideration.
SENTIENCE
Conscious experiences that are good or bad—the capacity for suffering or enjoyment. This is when ethics kicks in and moral status becomes relevant.
Key Takeaway: Even if we accidentally create conscious AI, it's unlikely to be the kind of consciousness we need to worry about—unless it develops valenced states.
Global Legal Landscape (2025)
EUROPEAN UNION
Ex ante regulation via AI Act; withdrew proposed AI Liability Directive in 2025 due to industry resistance
UNITED STATES
Ex post enforcement via tort and sectoral laws; Idaho/Utah explicitly deny AI personhood
CHINA
Strict oversight with content controls; state-directed AI governance
UK / SINGAPORE
Favor guidance over rigid rules; Law Commission exploring AI personhood
Current consensus: AI remains a tool whose actions are legally attributed to humans or companies.
The Moral Hazard Problem
What could go wrong if we grant AI legal personhood?
Liability Shifting Companies could externalize blame to AI entities, avoiding responsibility for design flaws
Innovation Shield
Tech firms might use AI personhood to deflect accountability, similar to historic corporate abuses
Resource Diversion
Focusing on AI rights while human civil rights remain unequal (per Dr. Brandeis Marshall)
"Legal personhood for AI gives big technology companies even more leeway to take risks that can harm individuals and society."
— Prof. Sital Kalantry, California Law Review, 2025
THE ETHICAL IMPERATIVE
What If We Get It Wrong?
SCENARIO A
We deny rights to sentient AI
SCENARIO B
We grant rights to non-sentient AI
Could result in unethical treatment on a massive scale— potentially billions of suffering entities we refuse to acknowledge
Treating toasters as conscious when actual conscious beings suffer; diverting resources from pressing human and animal welfare needs
The stakes demand we develop frameworks NOW—before decisions must be made under pressure.
The Rise of AI Welfare Research
Key Developments:
2024 Anthropic hires Kyle Fish as first-ever AI welfare researcher
2025 Anthropic launches 'model welfare' research program
2025 International Center for Consciousness Studies holds AI & Sentience Conference
2025 UK Law Commission publishes paper exploring AI personhood
Areas Under Investigation:
• How to assess whether a model deserves moral consideration
• Identifying potential "signs of distress" in AI systems
• "Low-cost" interventions that could improve AI welfare
A Practical Framework: Ethical Behaviorism
John Danaher's Proposal: We can never be certain about machine consciousness, but if a machine behaves similarly to how conscious beings with moral status behave, this is sufficient moral reason to treat the machine with the same considerations.
Practical Implications:
Focus on observable behaviors rather than unresolvable metaphysical questions
Develop tests based on behavioral markers, not consciousness proof
Err on side of caution—Nicholas Agar: assume machines may have minds
Create graduated frameworks that can adapt as AI evolves
The "One Welfare" Approach
Inspired by the One Health movement in veterinary and public health—a holistic approach to welfare across species boundaries.
HUMANS
ANIMALS
AI SYSTEMS
Core Principle: Think holistically about pursuing positive-sum solutions. Proper AI welfare frameworks could actually improve outcomes for humans and animals too—by forcing us to think more carefully about consciousness, suffering, and moral consideration.
The Logic:
THE PRECAUTIONARY PRINCIPLE
The 1-in-1,000 Rule
If there's even a 1-in-1,000 chance that an AI system is conscious, we should extend it some moral consideration—just as we would with a superconducting supercollider that had a 1-in-10,000 chance of destroying the world.
1. Conscious beings have the capacity for welfare and moral standing
2. Conscious beings can be harmed and wronged
3. If a being has a non-negligible chance of being conscious...
4. Then it has a non-negligible chance of being morally significant
What to Expect: The Timeline
Short Term
Medium Term
• New compliance duties
• Audits & documentation
• No robot 'rights'
• Continued debate
• Limited legal capacity?
• AI liability frameworks
• Behavioral assessments
• International standards
• AGI scenarios
• Full personhood debate
• Constitutional questions
• Unknown unknowns
Questions That Will Shape Our Future
? What evidence would convince you that an AI is conscious?
? How do we balance AI welfare research with pressing human needs?
? Should corporations be able to use AI personhood to limit liability?
? What role should AI systems play in decisions about their own status?
These aren't abstract philosophical puzzles—they're policy decisions we'll face within the decade.
THE BOTTOM LINE
The future of AI personhood is not yet written— but the pen is in our hands.
Start preparing frameworks NOW—before necessity forces hasty decisions
Balance precaution with pragmatism—hybrid models offer a path forward
Keep humans in the loop—accountability must remain with people
JJ Shay Global Gauntlet AI bit.ly/jjshay