legal personality for artificial intelligence? jacob turner personality for... ·...
TRANSCRIPT
Legal Personality for Artificial Intelligence?
Jacob Turner
Content
1. Legal and moral rights distinguished
2. Moral arguments for AI rights
3. Pragmatic arguments for AI legal personality
4. Challenges to AI legal personality
5. Helpful features for AI legal personality
(1) Legal and moral
rights distinguished
Rights for AI: Two Justifications
PragmaticEthical
Moral Rights
Legal Rights
(2) Moral arguments for AI
rights
Rights as Fictions: Animal Rights
UK Ill Treatment of Horses and Cattle Act 1822
Singapore Wild Animals and Birds Act Cap.
351 2000; Animals and Birds Act Cap. 7 2002
UK Animal Welfare (Sentencing and
Recognition of Sentience) Bill 2017
Naruto v Slater US Court of Appeals 9th Circuit.
Case: 16-15469, April 2018
Claim for habeas corpus rights for
chimpanzees rejected by New York Court of
Appeals, May 2018 ©David Slater…?
Rights as Fictions: Human Rights
Slavery - Dred Scott v Sandford (US
Supreme Court) 60 U.S. 393 1857
Universal Declaration of Human
Rights 1948
Gay marriage - Obergefell v Hodges,
(US Supreme Court) 576 U.S. 2015
Abortion – Irish Referendum 2018
Chief Justice Taney
Rights as Fictions: Environmental Rights
Christopher D. Stone, Should Trees
Have Standing? (1972)
Rome Statute of the International
Criminal Court (1998), Art. 8(2)(b)(iv)
Constitution of Ecuador (2008), Art.
56-60
Ganges and Yamuna Rivers granted
legal status by High Court of
Uttarakhand state in March 2017.
Overruled by Supreme Court in July
2017
“Methuselah”, Inyo County, California
Conclusions on Ethical Justifications for AI
Rights
Rights are not set in stone
If the recognition of rights depends on objective criteria, then we should extend rights to any entity which fulfils those criteria
Criteria which justify the grant of rights to other entities include the ability to suffer and compassion
If AI were to fulfil these criteria, then as a matter of ethical consistency we should extend similar protections to it
(3) Pragmatic Arguments
for AI Legal Personality
Legal Personality for AI
1.Is it possible?
2.Is it desirable?
Is Legal Personality for AI Possible?
Legal personality as a bundle of rights and obligations –Hans Kelsen, General Theory of Law and State (1945)
“A corporation is an artificial being, invisible, intangible, and existing only in contemplation of law. Being the mere creature of law, it possesses only those properties which the charter of its creation confers upon it, either expressly or as incidental to its very existence.”
- Chief Justice Marshall in Trustees of Dartmouth College v. Woodward 17 U.S. (4 Wheat.) 518 (1819).
Is Legal Personality for AI Possible?
Option 1: Housing AI within existingcorporate structures
Bayern,“The Implications of Modern Business-Entity Law for the Regulation of Autonomous Systems” European Journal of Risk Regulation (2) (2016), 297
Bayern, Burri, Grant, Häusermann, Möslein, and Williams, “Company Law and Autonomous Systems: A Blueprint for Lawyers, Entrepreneurs, and Regulators”, Hastings Science and Technology Law Journal, (2017) Vol. 9:2, 135-161
Likely effect of AI legal personality
Mutual recognition of (profit-making) legal entities in other EU States
“Companies or firms formed in accordance with the law of a MemberState and having their registered office, central administration orprincipal place of business within the Union shall, for the purposes ofthis Chapter, be treated in the same way as natural persons who arenationals of Member States.
‘Companies or firms’ means companies or firms constituted undercivil or commercial law, including cooperative societies, and otherlegal persons governed by public or private law, save for those whichare non-profit-making.”
- Art. 54, Treaty on the Functioning of the EU
Is Legal Personality for AI Possible?
Option 2: Housing AI within newcorporate structures
No closed list of possible legal persons
UK Court of Appeal recognises legal personality for Indian Temple which is “little more than a pile of stones” - Bumper Development Corp. v Commissioner of Police for the Metropolis [1991] 1 WLR 1362
Is Legal Personality for AI Possible?
Sophia the robot, ‘Citizen’ of Saudi Arabia
Is Legal Personality for AI desirable?
Justification 1: Bridging the “Accountability Gap”:
Who is responsible if AI causes harm?
Tort, contract, product liability, other strict liability? (see Module 2)
“For tomorrow’s agents… applying and extending existing doctrines in these ways may stretch legal interpretation to the point of breaking”
- Koops, Hildebrandt, Jaquet-Chiffell, ‘Bridging the Accountability Gap: Rights for New Entities in the Information Society?’ Minnesota Journal of Law, Science & Technology, 2010;11(2):497-561.
European Parliament’s Proposal
The European Parliament…
Calls on the Commission, when carrying out an impact assessment of its future
legislative instrument, to explore, analyse and consider the implications of all possible
legal solutions, such as ….
creating a specific legal status for robots in the long run, so that at least the most
sophisticated autonomous robots could be established as having the status of
electronic persons responsible for making good any damage they may cause, and
possibly applying electronic personality to cases where robots make autonomous
decisions or otherwise interact with third parties independently
- European Parliament resolution of 16 February 2017 with recommendations to
the Commission on Civil Law Rules on Robotics (2015/2103(INL)), 16 February
2017, (58 c.)
Is Legal Personality for AI desirable?
Justification 2: Bridging the “Retribution Gap”:
Difficulties identified in Module 2 of applying mens rearequirements to humans when crimes were immediated‘committed’ by AI
Avoids the “Retribution Gap” - see John Danaher, ‘Robots, Law and the Retribution Gap’ Ethics and Information Technology 18 (4):299–309 (2016)
Is Legal Personality for AI desirable?
Justification 3: Agents without Principals
AI now capable of concluding contracts. Raises questions as to agency and authority.
If AI agent was acting outside actual or ostensible authority then who is responsible for the obligations?
See further: Francisco Andrade, Paulo Novais, Jose Machado, and Jose Neves, ‘Contracting Agents: Legal Personality and Representation’, Artificial Intelligence and Law, Vol. 15 (2007), 357–373.
Is Legal Personality for AI desirable?
Justification 4: Distributing the fruits of AI ‘creativity’
Copyright, Patents, Designs
Free Speech
© Google (?)
Is Legal Personality for AI desirable?
Justification 5: Driving economic growth as part of a national AI strategy
Potential advantage to small, sophisticated, fast-moving nations.
Compare reforms of corporate law by “Offshore” jurisdictions to encourage companies to locate there
Consequent growth in local services industry, tax revenues, prestige, expertise
Ugland House, Grand Cayman
(4) Challenges for AI
Personality
Societal Disapproval
“We, Artificial Intelligence and Robotics experts, industry
leaders, law, medical and ethics experts… believe that
from an ethical and legal perspective, creating a legal
personality for a robot is inappropriate whatever the legal
status model”
- Open letter to the European Commission, April 2018
Response: such objections risk conflating moral and legal
personality, and thereby falling into the “Android Fallacy”.
Not all legal persons need the same rights.
AI Legal Personality as a ‘Liability Shield’
“We take the main case of the abuse of legal personality
to be this: natural persons using an artificial person to
shield themselves from the consequences of their
conduct”
- Joanna J. Bryson, Mihailis E. Diamantis, and Thomas D. Grant, ‘Of,
For, and By the People: The Legal Lacuna of Synthetic Persons’,
Artificial Intelligence and Law, Vol. 25, No. 3 (September 2017), 273–
291
Response: We can already pierce the corporate veil when there is
wrongdoing by an owner or a director of a company or other abuse of the
corporate structure. See Prest v Petrodel Resources Ltd [2013] UKSC 34,
and Goh Chan Peng v Beyonics Technology Ltd [2017] 2 SLR 592
AI rights being privileged over human rights
Fear that AI rights will override human rights
Coupled with issues of public unrest, concerns over
technological unemployment
Will AI rights rub salt into the wounds? Possibility of a
“tech-lash”, neo-Luddism
Response: As with all pragmatic exercises, the benefits
would need to outweigh the costs. We generally accept
that it is appropriate for company rights to outweigh natural
persons’ rights in some circumstances.
(5) Helpful Features for AI
Legal Personality
Identification of the AI
Legal persons generally represent unique, indivisible entities
AI systems may not follow this pattern – particularly if they
use ‘edge learning’ (i.e. localised versions of AI not
connected to the cloud) or utilise other types of distributed
networks
AI systems also constantly changing state
Potential Solution 1: Registration of AI source code. Use
distributed ledger technology (blockchain) to do so?
Potential Solution 2: AI could have parent and subsidiary
entities where systems are linked to each other
Ownership Trees
Parent Company
Division 1
Subsidiary
OpCo
Division 2 Division 3
Subsidiary
AI Rights
Likely to be restricted to economic ones in the first instance:
o Hold property (including IP it creates)
o Sue in Court
o Make contracts in their own name
Further civil rights which may be justified by circumstances
o Right to free speech/ expression (though perhaps within
tighter bounds than humans and corporations?)
o Compare: Citizens United v. Federal Election
Commission, 558 U.S. 310 (2010)
AI Obligations
Ability to incur liability, debt
Necessary in order for AI to “make good the damage it may
cause”
May meet some of the societal objections as to AI
unaccountability and use as a liability shield
Potential for AI criminal liability also? Compare corporate
criminal liability
What would the sanctions be? Fines? Deletion of AI system?
Recall that a company has “no body to kick and no soul to
damn”
Minimum Asset Requirements
Require that AI have a minimum level of assets in order to
operate in a particular environment (i.e. membership of a
trading room, or driving on a network of roads)
Compare: Basel III minimum regulatory capital requirements
for banks
Alternative: Compulsory insurance of AI systems - proposed
by European Parliament in February 2017 Resolution
Ownership and Management of AI
AI entity might be owned by
another AI entity
AI entities could then be owned by
a corporate person, or a human
(or any combination thereof)
AI entity might be managed by a
board of human directors
Possible (in theory) for a director of a company to be a non-natural
person
In 2014, an AI system was reportedly “appointed” to the board of
directors of a Hong Kong venture capital firm to assist in decision-
making.
Conclusions on AI Legal Personality
Rights can be moral, legal or both
Ethical justifications : suffering, instrumentalism, and
trans-humanism
Pragmatic justifications : accountability for harm caused,
ownership of benefits, and fostering innovation
Remaining challenges: identification, minimum asset
requirements, threshold for personality, control and
ownership