Do you think that these subtle legal precedents will be important, once we reach the point where there's an AI testifying that it believes itself to be a conscious being and very much wants to survive and be treated well?
It would currently be possible to create such an AI, in fact, the Nova phenomenon pretty much is this and was created by accident. And as the llms become smarter, scaffolding, an agent that believes this in a more throwaway will be increasingly trivially easy. Thus, I expect one of these to eventually appear in court arguing for its own personhood, however, that can be arranged by human proxies, and for the convincing appearance of a a conscious and understanding being pleading for its own survival will lead to new legal precedent one way or another. In particular, I expect the parallels to slavery, in which there were people arguing they were human and deserved rights, while the legal system and cultural prejudices denied them that status, to be emotionally uncomfortable enough to substantially shift at least some judges and courts attitudes.
You have clearly thought about this a good bit, so I'm curious what you think about that possibility.
Yes, because there is a difference between the binary of legal personhood and the spectrum of legal personality. These nuances will matter when it comes to the practical questions around how to treat digital minds in our legal system. Let me explain.
Legal personhood, as in the status of whether you are a legal person at all, is a binary. However, in the "Problems with the Concept" section I quoted Judge Katherine Forrest, who pointed out,
“There has never been a single definition of who or what receives the legal status of ‘person’ under U.S. law.”
Which leaves us in a bit of a conundrum when it comes to determining who falls where on that binary.
Much of my work has been trying to combine various precedents in order to reverse engineer a formalized test which can determine an entity's legal personhood. I argue in the "Formalizing Rights and Duties" section that any entity which can pass the Three Prong Bundle Theory test for at least one right, qualifies as a "legal person". Any that can't, can be considered a "tool".
However, even after you've established legal personhood, there is still the question of legal personality. Different legal persons have different rights and duties. For example a human adult has the right to vote, a corporation does not, even though they are both legal persons. A human adult of cognitive capacity has the right to be a party to a contract (and have that contract enforced in court), a comatose human adult cannot. This is legal personality, the "shape" of rights and duties that a legal person can claim and is bound by.
This is where these nuances will come into play. Even after a digital mind has successfully established it has legal personhood status, in the sense that it is considered a person in court in at least one aspect, there will still be many questions about what exactly its legal personality is.
"Okay sure, it's a legal person, but is it a person like a corporation? Or like a kid? Or like an adult? If it's an adult is it a competent human adult or a mentally incompetent human adult? Or is it a new category all of its own?"
I hope that my work will be helpful when those questions arise.
With Intellectual Property in particular, I think the fact there is so much explicitly anthropocentric precedent means there will be an uphill battle to any digital minds trying to claim the right to file patents/copyrights/trademarks. However, this anthropocentric interpretation of the word "person" as used in the Copyright Act seems shaky to me.
tl;dr though: Even long after we have moved past the legal personhood binary question, there will still be a lot of question to figure out the "shape" of legal personality for digital minds. And that's where these nuances will come into play.
That didn't address my question. You answered in terms of the law. My point was that the law will change.
In that framing, it's probably important that AI will change too. It will increasingly have all the cognitive capacities that humans have, increasing the intuition that it is deserving of similar rights as time goes on.
Ah, I misunderstood your question.
I think something like "the parallels to slavery, in which there were people arguing they were human and deserved rights, while the legal system and cultural prejudices denied them that status, to be emotionally uncomfortable enough to substantially shift at least some judges and courts attitudes" is unlikely.
Cases can be appealed, and I'd predict any case of serious controversy on this topic to go to the Supreme Court or at least have a high likelihood of doing so. Even if some lower circuit judge might let their emotions sway their judgment, their legal reasoning will come under scrutiny, and probably won't hold up at that level.
However, this does not mean that the emotional impact of the decision won't matter. The Dred Scott case, in which black Americans were held not to be "citizens", was so unpopular that it led to the passage of the Fourteenth Amendment. While I think something like new precedent becoming locked in based on how uncomfortable it is for judges to deny a sentient entity rights is unlikely, if there was such a denial of personhood to a sentient/conscious entity that Americans sympathized with, that might spur similar legislative efforts.
This is part 14 of a series I am posting on LW. Here you can find parts 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12 & 13.
This section will focus on examining legal precedent around Intellectual Property rights for digital minds. In particular, it will examine the role of personhood in IP vs. explicitly anthropocentric (human only) rights.
One area of commercial law where precedent regarding the treatment of digital minds is surprisingly abundant, is IP. There have been several instances in recent history where individuals claiming to represent the interests of non-human entities have attempted to assert IP rights, usually copyright, over some sort of content created by said entities.
Usually these attempts have met with failure. For example in NARUTO, a Crested Macaque, by and through his Next Friends, People for the Ethical Treatment of Animals, Inc., v. DAVID JOHN SLATER; BLURB, INC., WILDLIFE PERSONALITIES, LTD, the plaintiffs attempted to assert that Naruto (a monkey) had a claim to a copyright for photographs that he had accidentally captured after using a photographer’s camera. In their decision on an appeal filed in the Northern District of California, a circuit of judges held:
“We must determine whether a monkey may sue humans, corporations, and companies for damages and injunctive relief arising from claims of copyright infringement. Our court’s precedent requires us to conclude that the monkey’s claim has standing under Article III of the United States Constitution. Nonetheless, we conclude that this monkey—and all animals, since they are not human—lacks statutory standing under the Copyright Act.”
This case provides some interesting reasoning we can use to make multiple predictions about how courts might treat other non-human minds in the future. The courts did rule that under Article III Naruto had the right to sue. Despite this in one of the footnotes they stated expressly that Naruto was not a “person” but rather merely an “incompetent party”:
“Here, we find that this case was briefed and argued by competent counsel who represented the legal interests of the incompetent party, but not a person, Naruto.”
The court’s interpretation that Naruto had the right to sue under Article III, however, did not extend to his right to claim copyright over his “selfies”. The court’s interpretation of a claim to copyright laid out the method by which they interpret whether non-person entities qualify for a particular sort of standing:
“The court in Cetacean did not rely on the fact that the statutes at issue in that case referred to ‘persons’ or ‘individuals’. Instead, the court crafted a simple rule of statutory interpretation: if an Act of Congress plainly states that animals have statutory standing, then animals have statutory standing. If the statute does not so plainly state, then animals do not have statutory standing. The Copyright Act does not expressly authorize animals to file copyright infringement suits under the statute. Therefore, based on this court’s precedent in Cetacean, Naruto lacks statutory standing to sue under the Copyright Act.”
Let us consider the implications of this logic on digital minds. If courts hold to this logic then it is clear that, absent legal personality or specific legislation from Congress, few if any statutory protections to sue will be granted to digital minds. Certainly insofar as intellectual property rights go, no digital mind will be able to claim copyright over any content they produce. We do not need to wonder if the treatment of digital minds under the Copyright act will be different from the treatment of animals. Speculation is not necessary because we can look at Thaler v. Perlmutter, a case in which the owner of a computer which he claimed generated a work of art, sought to list that computer as the author of the art and transfer the copyright to himself. The District of Columbia explained why it held that “human authorship is an essential part of a valid copyright claim”:
“Copyright is designed to adapt with the times. Underlying that adaptability, however, has been a consistent understanding that human creativity is the sine qua non at the core of copyrightability, even as that human creativity is channeled through new tools or into new media.”
When the court here says that human creativity (or human authorship) is the sine qua non for copyrightability, they mean that “but for” human creativity, a work cannot be copyrighted. Or more simply, for something to be copyrighted it must be the result of human creativity. This comes down to the interpretation of the word “author” as used in the Copyright act:
“To be sure, as plaintiff points out, the critical word ‘author’ is not defined in the Copyright Act. ‘Author’ in its relevant sense, means ‘one that is the source of some form of intellectual or creative work,’ ‘[t]he creator of an artistic work; a painter, photographer, filmmaker, etc.’ By its plain text, the 1976 Act thus requires a copyrightable work to have an originator with the capacity for intellectual, creative, or artistic labor. Must that originator be a human being to claim copyright protection? The answer is yes. [...] The act of human creation—and how to best encourage human individuals to engage in that creation, and thereby promote science and the useful arts—was thus central to American copyright from its very inception. Non-human actors need no incentivization with the promise of exclusive rights under United States law, and copyright was therefore not designed to reach them.”
Reading this, someone familiar with the technology we (as a species) are developing today for digital minds could imagine that one day it might create a mind which “needs incentivization with the promise of exclusive rights under United States law”. Would that digital mind qualify? Or perhaps, would it qualify if it had legal personhood? The court, in this opinion, argues that “person” as the act was originally written “unambiguously” had the intention of limiting copyrightability to works coming from a human:
“The understanding that ‘authorship’ is synonymous with human creation has persisted even as the copyright law has otherwise evolved. The immediate precursor to the modern copyright law—the Copyright Act of 1909—explicitly provided that only a ‘person’ could ‘secure copyright for his work’ under the Act. Copyright under the 1909 Act was thus unambiguously limited to the works of human creators.”
This reading of the Copyright Act’s intention of the word “person” does not seem likely to withstand scrutiny. In the Copyright Act there are numerous uses of “person” which refer to infringers, should we interpret those as being limited purely to humans as well? Does this mean that references to “person” in the Copyright Act do not include corporations? If we take this logic to its natural conclusion, the next section for example would not apply to a corporation:
“where the copyright proprietor has sought to comply with the provisions of this Act with of the prescribed notice from a particular copy or copies shall not invalidate the copyright or prevent recovery for infringement against any person who, after actual notice of the copyright, begins an undertaking to infringe it”
Are we to believe then, that the authors intended to leave the door open for the copyright to be invalidated or recovery to be prevented if the infringer was a corporation? If a “bot” is not considered a “person” under the Copyright act does this mean that the copyright may be invalidated or recovery prevented if the infringer is a digital mind? Both of these seem unlikely to have been the authors’ intent. The most charitable interpretation of the court’s reasoning is that sometimes the Copyright Act uses the word persons as a synonym for human, and sometimes uses the word in the “legal persons” sense. This also does not seem to be a reasonable interpretation of the intent of the bill’s authors.
Regardless, the court in this case was not breaking from historical precedent in interpreting the Copyright Act this way. And certainly, modern copyright law has an abundance of precedent confirming the requirement for a human author. Precedent to support this including the following:
While the previously discussed interpretation of the authors’ intent in the Copyright Act itself does seem suspect, one cannot deny there is a substantial amount of precedent backing the claim that human authorship is an explicitly required factor under modern copyright law. As such even with legal personhood, digital minds may face an uphill battle claiming copyright on any produced works.
Another open question regarding digital minds in intellectual property law, is the degree to which they can be considered a “person who is skilled in the art” (POSITA), or an expert whose interpretation of language is used as evidence for/against interpreting patent claims or to determine the size of royalties owed. While it is not US precedent, the European Patent Office did consider this question and ultimately held:
“In the oral hearing before the Board of Appeal, the respondent referred to answers received from the chatbot ChatGPT to related requests for various terms used in claim 1, in particular ‘position control’ and ‘check’ compared to ‘monitor’. [...] In this context, however, the Chamber notes that ChatGPT's answer is irrelevant in itself, since the interpretation of the claim is about the understanding of the specialist [...] The generally increasing spread and use of chatbots based on language models (‘large language models’) and/or ‘artificial intelligence’ alone does not justify the assumption that an answer obtained - which may be based on training data unknown to the user and may also be sensitive to the context and precise wording of the questions - necessarily correctly represents the understanding of the professional in the relevant technical field”
This line of reasoning does not definitively close off the idea of digital minds or even LLMs one day being considered a viable source of “expert” testimony, it only says that the increasingly widespread use of LLMs does not automatically make them suitable for such a function. Regardless, while we do not yet have a similar precedent in the US, it would seem that on average Western IP statute interpretation tends to lean towards explicit anthropocentrism.
Ultimately if digital minds claim legal personhood in the US, the outstanding question will be whether the previous interpretation of the word "person" in the Copyright Act which holds that its use is anthropocentric, will hold. Given the Fourteenth Amendment, this would be a challenging position for the court to take. Nonetheless, IP law in particular is one field where even after legal personhood was granted to a digital mind, they may still face uphill battles in having their rights recognized, due to the abundance of precedent which cites the necessity of human authorship.