The Legal Status Of AI As A Juridical Person - A Step Too Far
Author: LawSphereX
Category: Technology, Cyber & Privacy
Published:
Updated:
THE LEGAL STATUS OF AI AS A JURIDICAL PERSON: A
STEP TOO FAR?
Mrunalini K Sonkusare, OP Jindal Global University
ABSTRACT
The question of whether advanced artificial intelligence (AI) systems should ever be treated as legal persons has recently entered policy debates, fueled by suggestions such as the European Parliament’s 2017 call for examining “electronic personhood” for autonomous1. This article argues that extending juridical personhood to AI is both legally unwarranted and pragmatically dangerous. It surveys the concept of legal personhood – historically limited to natural persons and human-created entities (e. g. corporations) – and shows that AI lacks the defining features underlying existing legal personhood. Legal analyses highlight how granting personhood to machines would undermine accountability: AI “persons” could become entities that are either “accountable but unfunded” or “fully financed but unaccountable,” weakening protection for actual human victims.
Ethical arguments emphasize that AI lacks moral agency and consciousness, making any legal rights or duties conferred on them incoherent. The comparative section reviews multiple legal systems (including the US, EU, Japan, China, and others) and finds that none have adopted AI personhood; many have even barred it (e. g. Utah’s new law) while focusing on liability or safety regulations instead. In sum, scholars and legislators are widely skeptical of the idea: leading experts have urged that the very proposal of “electronic person” status be “discarded from both a technical perspective and a normative viewpoint”. 2 The article concludes that existing legal frameworks, suitably adapted, can address the challenges posed by AI, and that creating a new class of legal person for machines would be premature and perilous.
1 Liisa Janssens, ‘A Prospect of the Future: How Autonomous Systems May Qualify as Legal Persons’ in Emre Bayamlıoğlu and others (eds), Being Profiled: Cogitas Ergo Sum – 10 Years of Profiling the European Citizen (Amsterdam University Press 2018) https://www. jstor. org/stable/j. ctvhrd092. 24 accessed 25 June 2025 2 National Conference of State Legislatures, Report: Legislation Related to Artificial Intelligence (NCSL, 2023) https://www. ncsl. org/technology-and-communication/legislation-related-to-artificial-intelligence accessed 28 June 2025
Introduction
Legal personhood has traditionally been reserved for human beings and certain artificial legal entities like corporations, which are created by law to serve human economic purposes. By contrast, AI systems are manufactured artifacts. Recently, however, the rapid development of autonomous AI has prompted debate on whether highly advanced AI might one day occupy a legal status akin to a juridical person. This idea gained prominence in 2017 when the European Parliament’s Resolution on Civil Law Rules on Robotics invited the Commission to explore granting “electronic personhood” to robots that make autonomous decisions3. News of a humanoid robot (“Sophia”) receiving honorary citizenship in Saudi Arabia also heightened the intrigue. In this context, it is crucial to ask: what does legal personhood mean, and should it ever be extended to machines?
As one scholar observes, a legal person simply “has the capacity to sue and be sued, to hold property in its own name, and to enter into contracts”4. These capacities are not inherent features of the entity itself but are conferred by law for practical reasons. Even among legal persons, rights and powers are carefully circumscribed by law and by constitutional principles. For instance, corporations in law can hold property and enter contracts, but they cannot marry or vote, and they can only act within the scope of their charter5. Thus, granting personhood is a conscious choice to endow an entity with rights and obligations in order to accomplish specific social goals. Whether AI systems belong in that category is the question. This article argues that it is excessive to give AI legal personhood. A brief description of legal personhood is given in Section I.
Legal objections to AI personhood are discussed in Section II, with an emphasis on doctrinal, liability, and accountability concerns. Examining ethical and philosophical issues, Section III highlights AI's lack of moral agency and true autonomy. The methods of several jurisdictions, including the US, EU, Japan, and China, are compared in Section IV. It reveals that none have accepted AI personhood, and many are openly opposing 3 Liisa Janssens, ‘A Prospect of the Future: How Autonomous Systems May Qualify as Legal Persons’ in Emre Bayamlıoğlu and others (eds), Being Profiled: Cogitas Ergo Sum – 10 Years of Profiling the European Citizen (Amsterdam University Press 2018) https://www. jstor. org/stable/j. ctvhrd092. 24 accessed 25 June 2025
4 Samir Chopra and Laurence F White, A Legal Theory for Autonomous Artificial Agents (University of
Michigan Press 2011)
5 Samir Chopra and Laurence F White, A Legal Theory for Autonomous Artificial Agents (University of
Michigan Press 2011) it. The conclusion makes the case that existing liability regimes and regulations are a better way to handle the issues raised by AI than is establishing a new class of legal person. I. The Nature of Legal Personhood Legally, a "person" does not have to be a human. In order to ease transactions and obligations, legal systems have long regarded governments, corporations, partnerships, and other entities as juridical persons. However, this synthetic personhood is precisely defined. The ability to carry out legal actions (contract, own property, sue and be sued) confers legal personhood, as explained by Chopra and White. However, this status is entirely dependent on the law; other legal persons only have the rights and powers that the law grants them, while only a “free human of majority age” is automatically entitled to the full spectrum of rights.
As artificial persons, corporations, for instance, are able to enter into business contracts and own property, but they are not allowed to get married or cast ballots, and their operations are restricted by statutes or charters. Legal personhood is therefore a legal construct that is suited to human interests rather than an innate trait. An entity is typically granted personhood for pragmatic reasons, like guaranteeing accountability. Think about a corporation: it has the ability to sue and be sued, giving it and its owners legal avenues for resolving conflicts. Importantly, a corporation must also have assets or money to support it, according to the law. A company must be able to pay damages if it is found to be liable in a tort.
Likewise, states and municipalities, which are also considered legal entities, may be sued for negligence, but only because they theoretically have the resources and budgets to be held accountable. To put it succinctly, personhood only makes sense if the entity is capable of carrying obligations and consequences. Given this background, extending personhood should be done with caution. What responsibilities and rights would AI systems have, and how would they fulfill them, if they were given legal personality? Only when a human (the AI machine's operator, an enclosing trust, or a corporation) is put on the hook can the machine enter into a contract or own property. If not, it would be a self-governing entity without accountability.
According to one legal scholar, every existing legal person has a link to human control or support—corporations have directors and shareholders, partnerships have partners—so that someone can enforce duties. Without a human owner, a robot would be a legal black hole.
Finally, conferral of personhood is a sensitive legal choice. As Nosworthy (cited by Chopra & White) notes, whether an entity deserves legal personhood is an evaluation of social importance that must consider its impact on legal and political institutions. 6Extending personhood to AI could fundamentally alter the balance of rights. For instance, the natural person model cannot be applied to a robot without absurd consequences: it would imply robots holding human fundamental rights (e. g. dignity, free movement, citizenship), which would conflict with constitutional guarantees for humans. 7 On the other hand, the corporate model assumes that people support the organization. Additionally, an AI "person" without human directors would violate that model. In other words, AI does not neatly fit the current models of legal personhood.
This conceptual misalignment portends more serious issues that will be covered later. II. Legal Arguments Against AI Personhood A. Accountability and Liability Liability is a key justification for personhood, which makes sure that someone is held accountable for damage. One could argue that autonomous AI systems, such as industrial robots or self-driving cars, could be held accountable for mishaps. In reality, though, legal systems forbid fining or punishing an entity that lacks conscious will or financial resources. An AI "person" that is sued needs to have assets in order to pay judgments. Any financial gain an AI makes (for example, through conducting business) would actually come from its creators or owners, according to the majority of proposals.
Therefore, giving the AI a "person" could just protect the people who created it by treating them like stockholders in an anonymous company. As Liisa Janssens reports, even supporters of electronic personhood recognize serious risks. Citing Bryson et al. (2017), she notes that creating “novel legal persons” (like an AI) has historically led to problems. One concern is what Bryson calls the “accountable but unfunded” case: the entity can be held responsible in theory, but has no money to pay damages. The opposite situation, “fully financed but unaccountable,” can also arise (for example, if an AI has
6 Samir Chopra and Laurence F White, A Legal Theory for Autonomous Artificial Agents (University of
Michigan Press 2011) 7 Ugo Pagallo, ‘Apples, Oranges, Robots: Four Misunderstandings in Today’s Debate on the Legal Status of AI Systems’ (2018) 376 Philosophical Transactions of the Royal Society A assets but no one responsible for its misconduct). In either case, the protection of human victims is weakened. In short, legal personhood for AI risks producing precisely the loopholes it was meant to close. 8 In contrast, human beings (manufacturers, owners, and operators) can be held accountable by our current liability systems. The makers or operators of machines may be held strictly or negligently liable under tort law. When defects cause harm, product liability regimes enable injured parties to sue manufacturers.
For instance, if a pedestrian is hurt by a robot car, the carmaker or software developer may be sued by the courts for negligence or product defects, and they may be required to have insurance or compensation funds. In a similar vein, employers may be held vicariously accountable for injuries brought on by machinery they manage. Without establishing new legal categories, these mechanisms can be reinforced (for example, by requiring liability insurance for AI manufacturers). Importantly, existing personhood for corporations already offers a useful analogy: when a corporation harms others, it can be sued as an entity and made to pay from its capital. But crucially, a corporation must have its own funds and is directly controlled by people. If we treated an AI as a corporation, we would need to designate human “directors” or require AI to own assets.
Without such links, AI-personhood simply shifts blame into the void. Indeed, Bryson warned that unless carefully designed, “the creation of legal personhood for electronic agents could allow individuals to avoid accountability by attributing actions to the robot instead of to themselves”. 9 In conclusion, assigning liability does not require the legal-person solution, and it may even make it less effective. The motivation was stated by the European Parliament itself: more autonomous robots could disrupt the traditional chain of liability, "making ordinary rules on liability insufficient. " However, it is a blunt tool to create a new category of personhood.
Regulators can impose stricter liability regulations in place of people, or they can mandate that any "smart bot" be supported by mandatory insurance (as some commentators have suggested) 8 Liisa Janssens, ‘A Prospect of the Future: How Autonomous Systems May Qualify as Legal Persons’ in Emre Bayamlıoğlu and others (eds), Being Profiled: Cogitas Ergo Sum – 10 Years of Profiling the European Citizen (Amsterdam University Press 2018) https://www. jstor. org/stable/j. ctvhrd092. 24 accessed 25 June 2025 9 Liisa Janssens, ‘A Prospect of the Future: How Autonomous Systems May Qualify as Legal Persons’ in Emre Bayamlıoğlu and others (eds), Being Profiled: Cogitas Ergo Sum – 10 Years of Profiling the European Citizen (Amsterdam University Press 2018) https://www. jstor. org/stable/j. ctvhrd092. 24 accessed 25 June 2025 or a manufacturer liability regime.
These methods maintain that people are ultimately responsible. B. Doctrinal and Constitutional Limits Even aside from liability, granting AI legal personhood could conflict with fundamental legal principles. The open letter by AI experts warned that if robots were made persons under naturalperson law, they would be entitled to human rights – an outcome “in contradiction with the Charter of Fundamental Rights of the EU and the Convention for the Protection of Human Rights”. For example, a robot “citizen” in Saudi Arabia raised eyebrows because it implied granting a mechanical being privileges reserved for humans. While Sophia’s honorary citizenship was largely symbolic, it illustrates the conceptual clash. As one commentator quipped, “Your cat makes autonomous decisions, too, but we do not hold the cat legally responsible for its actions”.
10 Conflating machines with people could undermine the foundation of rights from a constitutional standpoint. Human abilities (self-awareness, moral comprehension) that machines do not possess are the foundation of human rights and dignity. We run the risk of weakening the human-centered justification for certain rights if we grant a robot, for example, the right to free speech or the protection of personal data. Similarly, corporate personhood functions because it is based on the individuals who form the corporation; an AI-person would not have the same foundation. As Pandolfo points out, comparisons to slavery are more appropriate: one should make sure that third parties are protected when interacting with machines and that humans (masters) are not harmed by a robot's actions.
In other words, rather than treating robots as autonomous beings with rights, the law should view them as instruments that are supervised by humans. Furthermore, there isn't a legal classification for sentient machines yet. The categories upon which law is based—natural persons, corporations, trusts, governments, etc. —all have human or collective roots. Establishing a new classification for AI would lead to instability. Any changes must be made extremely carefully because legal personhood is a cliff, not a slippery slope. Prominent legal scholars contend that modern AI is very different from anything that 10 Ugo Pagallo, ‘Apples, Oranges, Robots: Four Misunderstandings in Today’s Debate on the Legal Status of AI Systems’ (2018) 376 Philosophical Transactions of the Royal Society, resembles a morally responsible, intentional agent.
Even though AI may eventually develop into intentional agents, Chopra and White acknowledge that they are currently "not moral persons. " They stress that personhood should only be granted when policy justifies it; this is a "pragmatic" choice that isn't given out by the mere existence of cutting-edge technology. 11 C. Practical Obstacles In practice, there are numerous logistical issues that arise even if we choose to attempt treating AI as people. How, for example, might an AI really interact with the law? A contract needs to show consent; since the majority of AIs nowadays lack will or comprehension, their "consent" is meaningless. In actuality, any AI contract or "ownership" would mirror software programming or human instructions.
Comparably, identifying an AI's assets would be necessary to enforce judgments against it; however, an AI only "has" what its human operator programs it with, such as owner-held bank accounts. It is still necessary to hold the people or businesses that created the AI "person" accountable if they are required to compensate damages. Accountability is another pragmatic issue. Could an AI face criminal penalties if it is a legal person? Obviously not through fines or incarceration, as it has no personal stakes. Fining an AI would entail taking away "its" money, which again is human property. It would be ridiculous to imprison an AI. The only possible punishment could be the machine being destroyed or rebooted, which would essentially shut it down. However, that hardly constitutes justice; it is the same as destroying evidence.
These contradictions imply that AI's legal personhood is unable to serve the purposes of conventional liability. At a minimum, proponents who suggest AI personhood implicitly propose heavy regulation and insurance schemes. The European Parliament’s resolution proposed setting up a compensation fund and compulsory insurance for robots, acknowledging that strict liability for AI could be placed on the “electronic person” itself or its owners. 12 Such schemes, however, can be put into place without establishing a new personhood category, as observers point out: AI developers or users can just be required to provide insurance and funds. The notion that
11 Samir Chopra and Laurence F White, A Legal Theory for Autonomous Artificial Agents (University of
Michigan Press 2011) 12 Liisa Janssens, ‘A Prospect of the Future: How Autonomous Systems May Qualify as Legal Persons’ in Emre Bayamlıoğlu and others (eds), Being Profiled: Cogitas Ergo Sum – 10 Years of Profiling the European Citizen (Amsterdam University Press 2018) https://www. jstor. org/stable/j. ctvhrd092. 24 accessed 25 June 2025 robots could self-insure by amassing wealth and then paying taxes on it, for instance, is speculative; in practice, any wealth is owned by underlying users or corporations. While requiring insurance for "smart bots" makes sense, it does not necessitate that they be legal persons; rather, it calls for insurance policies that are backed by humans. In summary, personhood is not specifically necessary for any of the practical arguments (insurability, liability pooling, and ease of suing). Existing legal tools can be modified to address all of these issues.
The development of AI as legal persons, according to one analysis, would not end the liability controversy but rather make it more complex, leaving people still intertwined with the machines. Global trends highlight this point: laws prohibiting the granting of personhood to non-human entities have already been passed in a number of jurisdictions. In reality, countries are taking the opposite course, shielding people from pointless attempts to achieve "electronic personhood" instead of advancing it. III. Ethical and Philosophical Concerns Beyond legal nuances, thinking of AI as persons raises serious ethical concerns. The first is the absence of moral agency. Having the capacity for moral comprehension or intentionality is frequently associated with being a legal person. AI systems function without consciousness, desires, or moral awareness both now and in the future.
They abide by learning rules and algorithms. It is morally acceptable to give a being legal rights or obligations only if the being has self-interest. A robot cannot feel pain, has no experiences, and no welfare interests. Therefore, it makes no sense to grant it rights; what would those rights safeguard? In a similar vein, obligations only make sense if the agent is able to understand and fulfill them. Since AI lacks the psychology to change, punishing it for misbehavior would neither teach it nor prevent further harm. Ethicists also note the anthropomorphic allure: humans often treat machines as if they were alive (the “media equation”), but this does not imbue them with actual personhood.
The renowned robot ethicist Kate Darling has famously said that granting robots personhood at this stage is “ideological and non-sensical” because it misunderstands both the nature of robots and the problem of liability13. It would essentially absolve humans of responsibility by shifting the 13 Charlie Stross, Experts Sign Open Letter Slamming Europe’s Proposal to Grant Robots Personhood Gizmodo (online, 17 January 2018) https://gizmodo. com/experts-sign-open-letter-slamming-europe-s-proposal-to-
1825240003 accessed 28 June 2025.
moral significance of actions from human actors to machines. For example, blaming the machine for a harmful act committed by a self-learning algorithm could absolve the developers. It is unethical to have this incentive structure inverted. Furthermore, granting AI legal rights could erode human-centric notions of dignity and equality. The idea of extending fundamental rights (e. g. life, liberty) to machines is typically rejected by human rights scholars. As the open letter warns, if robots were given legal status on the model of natural persons, they would hold rights that might conflict with humans’ rights14. Robots could assert rights that divert attention from human needs even in a corporation-like framework. Could an AI "person" demand, for instance, labor compensation or privacy protections? Such claims would be ridiculous because humans design and operate robots.
Thus, the ethical foundation of rights as safeguards for entities with subjective interests is compromised by personhood for AI. Finally, there are serious philosophical questions. Assumptions concerning consciousness, autonomy, and community membership are inherent in the concept of personhood. Giving AI personhood solely for practical purposes makes it a legal fiction, similar to giving a river or mountain human status (rights-of-nature debates). A lot of researchers warn that these comparisons are deceptive. People like Liisa Janssens contend that personhood should not be taken lightly and that AI shouldn't be grouped with "natural" or "legal" persons without taking into account the qualitative distinctions between them. Declaring an AI to be a "person" essentially distorts our moral lexicon, even though it may have a short-term benefit (liability).
Personhood must not be trivialized or politicized in order for legal policy to be sound. As Pagallo points out, rather than generating new categories of personhood based on speculation, the discussion should move forward with "pragmatic" legal solutions. In conclusion, the ethical component supports the legal arguments: giving AI personhood runs the risk of undermining human values because they lack the necessary moral consideration or accountability skills. As a result, the idea has been unanimously rejected by leading experts. The creation of an "electronic 'person'" status for robots should be discarded from both a technical perspective and a normative (legal and ethical) viewpoint," according to an open letter to the European Commission signed by more than 150 experts in AI and robotics.
The 14 Ugo Pagallo, ‘Apples, Oranges, Robots: Four Misunderstandings in Today’s Debate on the Legal Status of AI Systems’ (2018) 376 Philosophical Transactions of the Royal Society A scholarly consensus that the proposal is viewed as premature and conceptually flawed on its face is encapsulated in this strong language. 15 IV. Comparative Legal Analysis Though different jurisdictions have taken different approaches to AI regulation and liability, none have embraced the idea of granting legal personhood to AI systems. Rather, the majority of legal systems depend on expanding or modifying pre-existing structures. United States. American law currently treats machines as property, not persons. Neither federal nor state statutes grant AI any legal personhood. In fact, state legislatures have begun moving in the opposite direction.
For example, Utah enacted the Legal Personhood Amendments in 2023, which explicitly prohibit any governmental entity from granting legal personhood to “certain categories of nonhumans,” presumably including AI. 16In other places, similar bills have been proposed. Though they have consistently rejected them, U. S. courts have occasionally considered metaphorical arguments (such as petitions to treat an AI or animal as a person). Under product liability, negligence, or agency principles, manufacturers, programmers, or owners are liable for harm caused by AI in the United States. Tools to hold humans accountable are provided by common law doctrines like strict liability for abnormally dangerous activities. Instead of rethinking personhood, regulatory efforts concentrate on specialized fields (such as autonomous vehicle safety standards).
To put it briefly, the United States views AI less as sui juris entities and more as regulated machines or tools. European Union. The EU has been the most visible forum for the personhood debate, but it remains unimplemented. The 2017 EU Parliament resolution (non-binding) explored the idea of an “electronic person” for sophisticated robots to ensure liability for damage. 17However, that proposal was heavily criticized. The expert open letter (referenced above) was signed by EU and international scholars, companies and even roboticists.
Subsequently, the European Commission’s own experts concluded that existing law could be adapted and that formal 15 Ugo Pagallo, ‘Apples, Oranges, Robots: Four Misunderstandings in Today’s Debate on the Legal Status of AI Systems’ (2018) 376 Philosophical Transactions of the Royal Society A 16 National Conference of State Legislatures, Report: Legislation Related to Artificial Intelligence (NCSL, 2023) https://www. ncsl. org/technology-and-communication/legislation-related-to-artificial-intelligence accessed 28 June 2025 17 Liisa Janssens, ‘A Prospect of the Future: How Autonomous Systems May Qualify as Legal Persons’ in Emre Bayamlioğlu and others (eds), Being Profiled: Cogitas Ergo Sum (Amsterdam University Press 2018) personhood was unnecessary. EU policy has instead moved towards the proposed AI Act and liability reforms.
The draft AI Act (as of 2025) focuses on risk categories and human oversight, with no mention of personhood. Similarly, a 2019 EU Parliament-commissioned study on civil liability acknowledged the concept but recommended improving product liability and insurance schemes rather than creating new legal subjects. There is also opposition among EU member states. Experts from France and Italy have also stressed human liability, while Germany's robotics strategy and regulatory statements have specifically minimized any intention to give robots legal status. Conversely, some nations with civil law have thought about complementary measures. For instance, a resolution passed by the EU Parliament recommended mandatory insurance pools and a European-wide registry of sophisticated robots.
These plans seek to provide victims with compensation, but they would not rely on the robots themselves but rather on human contributions to the funds. No EU nation has enacted legislation acknowledging robots as human beings. Similar to any other dangerous tool or new product category, AI is generally being treated in Europe under the current civil liability framework. In fact, certain scholars point out that giving AI rights might be in opposition to EU human rights legislation, supporting the idea that robots shouldn't be considered human. 18 United Kingdom. The UK, although no longer an EU member, has followed a similar path. In
2017 the House of Lords conducted inquiries on AI ethics and law. The committee even posed
the question in hearings: “Does the creation of some form of electronic personhood need to be considered in the UK? ”. However, that strategy hasn't been used in later reports or policies. Assigning responsibility to human designers, operators, or owners is still the focus of UK law. For example, the UK Centre for Data Ethics's tech guidance promotes accountability and transparency measures for AI systems without recommending that the algorithms themselves have legal standing. Because of this, the UK shares the same viewpoint as the US: machines are machines, and the people who operate them are still accountable. Japan, a leader in robotics, has also rejected legal personhood for AI. A high-profile case illustrates this: in 2017 a municipal government granted a “special residence permit” to an AI chatbot named Shibuya Mirai, ostensibly treating it as a foreign dignitary.
This gesture was 18 Ugo Pagallo, ‘Apples, Oranges, Robots: Four Misunderstandings in Today’s Debate on the Legal Status of AI Systems’ (2018) 376 Philosophical Transactions of the Royal Society A quickly criticized for having no legal basis; it contradicted Japanese immigration law, which only applies to human foreignersersj. eu. In other words, even Japan’s symbolic experiment underscored that existing laws don’t count AI as legal subjects. Japanese regulators focus on safety and testing zones (e. g. special “regulatory sandboxes”) for robotics, not on any personhood status. The government’s recent AI strategy emphasizes standards, privacy safeguards, and industrial promotion, without hinting at autonomous legal rights for machines. Indeed, legal scholars have noted that Japanese legislation on robotics has “satisfactorily regulated the use of AI... while denying it legal personality”.
China likewise has not moved toward personhood for AI. Instead, Chinese policy has been to create guidelines and technical standards for robotics. In July 2024, Shanghai authorities issued governance guidelines for humanoid robots, requiring that they “not threaten human security” and must “effectively safeguard human dignity”. These rules reflect a human-centric approach: robots are to serve society under control, not to become independent entities. National AI development plans (such as the 2017 “Next Generation AI Development Plan”) set innovation targets but impose no legal person status. In practice, liability in China for machine-made harms is handled under existing laws: product liability applies to goods with AI components, and in tort the manufacturer or user is blamed for negligence. There is no Chinese concept of an AI legal person, and no legislative movement toward it.
19 In most other countries the trend is similar. Common law systems (Canada, Australia) have not contemplated AI personhood. Civil law countries in South America or Asia are mostly focused on data protection and cybersecurity rather than personhood. Some countries are exploring liability funds or insurance mandates for autonomous vehicles, but always in a way that ties the machine back to human accountability. No government has so far enacted a law recognizing AI systems as legal persons with rights and duties. Instead, many are explicitly prohibiting it: for example, Utah’s law echoes a broader “rights-of-nature” movement by barring nonhumans from personhood. This global stance indicates a widespread consensus: law makers recognize the impulse to fill liability gaps, but overwhelmingly conclude that creating AI persons is neither necessary nor wise.
20 19 Xinmei Shen, ‘China’s Laws of Robotics: Shanghai publishes first humanoid robot guidelines’ South China Morning Post (online, 7 July 2024) https://www. scmp. com/tech/policy/article/3269500/chinas-laws-roboticsshanghai-publishes-first-humanoid-robot-guidelines accessed 28 June 2025 20 Earth Law Center, ‘Utah Advances Anti-Rights of Nature Bill with Implications for Artificial Intelligence’ Earth Law Center (blog post, 28 February 2024)
Conclusion
The idea of granting artificial intelligence systems the status of juridical persons has attracted attention, but careful analysis shows it to be a step too far. Historically and conceptually, legal personhood has been a tool to order human and corporate relations, not a ready category for machines. Extending it to AI would break with the foundations of our legal system. As has been emphasized by scholars and even the EU open letter of experts, robot “personhood” must be discarded as “ideological and non-sensical”. 21 Legalizing AI would have disastrous real-world repercussions. It would lead to accountability gaps because victims would have no recourse if an AI with personhood possessed assets but lacked the consciousness to manage them.
On the other hand, requiring all AIs to be self-funded or insured would, in reality, leave all responsibilities to human backers, rendering the personhood myth unnecessary. Critics warn against precisely this situation: "entities that are accountable but unfunded, or fully financed but unaccountable" erode human rights under the law. To put it briefly, personhood would create more injustice and uncertainty than it would alleviate. AI is not a moral being that merits rights or responsibilities, according to ethics. They are devoid of the qualities that constitute personhood—consciousness, intent, and dignity. The line between creator and creation would be blurred if they were given even the minimal status of a corporation. Prominent AI ethicists contend that doing so would encourage producers to transfer accountability to machines, thereby relieving us of our moral obligation.
These claims are supported by comparative experience. All of the main legal systems in the world, including common law (US, UK), civil law (EU, Japan), and others (China), have consistently avoided giving artificial intelligence any unique legal personality. Conversely, some have passed laws that forbid it (such as the one in Utah). Rather, every jurisdiction tackles AI issues by expanding the regulations that already exist, such as revising tort and contract law, enforcing safety regulations, and in certain situations, requiring AI liability insurance or compensation funds. According to one Chinese guideline, robots must "protect human dignity," https://www. earthlawcenter.
org/blog-entries/2024/2/utah-advances-anti-rights-of-nature-bill-with-implications-f or-ai accessed 28 June 2025 21 Ugo Pagallo, ‘Apples, Oranges, Robots: Four Misunderstandings in Today’s Debate on the Legal Status of AI Systems’ (2018) 376 Philosophical Transactions of the Royal Society but that does not entail giving them legal rights. Everyone agrees that the most well-known AI "citizenship" case—Saudi Arabia's Sophia—is a publicity gimmick rather than a precedent. In conclusion, giving AI systems legal personhood is risky and unnecessary from a legal standpoint. Instead of creating a new category of "electronic person," it would be better to modify our laws in specific ways, such as requiring transparency, strengthening human liability, and perhaps even developing sui generis liability schemes.
AI should stay in the legal category of tools or products until it can actually comprehend and respect norms, which is not likely to happen anytime soon. It would be unfair to treat machines like people since it would obfuscate important differences. Because of these factors, it should be acknowledged that giving AI legal personality is excessive and goes against both good legal principles and moral discernment.