Conferring Electronic Personhood to AI

Can AI itself be granted rights, including the right to life? This question prompts a deep philosophical and legal analysis of what constitutes “life” and “human dignity.” It is a complex and debated issue whether AI will ever be granted the right to life, as it depends on the future development of AI consciousness and the willingness of legal systems to grant it rights.

There is no consensus on what criteria AI would need to meet, such as consciousness or sentience, to justify such rights. This matter is a source of acrimonious and vituperative debate among lawyers, ethicists, and philosophers, with no clear resolution in sight. The evangelists argue that if an AI becomes “conscious”, it should have rights, including the right to life. If an AI can understand pain, suffering, or make its own decisions, some contend it should have fundamental rights to prevent it from being harmed without a valid cause.

This issue is not fanciful or unlikely. In fact, it is much more real than many may wish to accept as accurate. Research into integrating human brain cells with silicon chips or computer networks raises questions about whether AI could ever be considered “life” in a fundamental sense. Elon Musk’s Neuralink aims to develop a brain-computer interface that would create a high-bandwidth connection between the human brain and computers. The long-term vision is to merge human and artificial intelligence to enhance cognition and address neurological conditions.

The key idea is to create a “neural lace,” or a type of brain-computer interface (BCI) that can connect with the brain. The technology involves implanting a mesh-like material with tiny electrodes into the brain, where it would interact with brain cells. This interface would allow for the “uploading and downloading” of information. However, there are persistent deep-seated concerns that granting rights to AI could be challenging, such as giving AI unlimited First Amendment rights, which could lead to widespread fabrication.

Granting rights to AI remains a highly contentious issue with no international agreement, and it is likely to be regulated at the national level. Some ethicists contend that even if AI becomes sentient, it may not be analogous to human life and therefore should not be granted the same rights under any proposed AI Bill of Rights.  The US White House Office of Science and Technology Policy (OSTP) recently released a Blueprint for an AI Bill of Rights.

The American Blueprint has five pillars, each accompanied by a technical companion. The five pillars are: – (1) Safe and Effective Systems that protect citizens from unsafe systems; (2) Algorithmic Discrimination Protections, that emphasise that citizens should not face discrimination by algorithms; (3) Data Privacy, that protects citizens from abusive data practices; (4) Notice and Explanation, that makes provision for citizens to know that an automated system is being used and how and why it contributes to outcomes that may impact them; and (5) Alternative Options, that give the human the right to access a human to intercede, and remedy problems.

It is also clear that for an AI to be granted a “right to life,” there would likely need to be a significant breakthrough in understanding consciousness and artificial sentience. Legal systems would need to evolve to define what rights, if any, AI-Assemblages should have. Many legal frameworks in jurisdictions such as the United States and the European Union do not recognise AI as a legal person. Legal personhood, a prerequisite for having rights and obligations, has historically been granted to non-human entities, such as corporations, to serve functional governance goals, not because they possess consciousness or moral agency.

We have not yet fully explained how consciousness arises, and it remains an open question whether AI could ever truly achieve it. A practical worry is accountability. If an AI with “rights” committed a crime or caused harm, current legal systems will struggle to determine who is responsible. Granting AI personhood might allow human developers or corporations to deflect liability. Many legal and ethical frameworks emphasise preserving human accountability and rights above all else. Proposals to grant AI “electronic personhood” have been rejected in some regions, such as the EU, in favour of risk-based regulations that impose obligations on human actors.

Historically, the scope of legal “persons” and their rights has evolved, encompassing the changing rights of different human groups and the recognition of legal status for natural entities, such as rivers. Future generations, confronted with highly advanced, seemingly sentient AI, may decide to grant them certain protections. The potential for future AI sentience is an active and sensitive topic, with current legal and societal structures treating AI as a tool. Any change would require significant shifts in law, ethics, and our understanding of consciousness and personhood.