Intellectual Property vs Artificial Intelligence
What is the Artificial Intelligence?
Artificial Intelligence, for Stuart Russell and Peter Norvig, authors of “Artificial Intelligence: A Modern Approach, Global Edition”, means a “field of studies in which intelligent agents are designed and built”. In what sense? What would be the etymological meaning of the term? “Agent” is “anything that can demonstrate the capacity to perceive one’s own environment through sensors, and to act on it through actuators”; “intelligent” identifies a “rational” agent, which “for each possible sequence of perceptions is able to select, among the various possible actions in response, the one that is presumed to maximize the measure of its performance, on the basis of the data deriving from the sequence of the same perceptions and the form of knowledge inserted within it”. It is also true that what is more incisive in this binomial, is the “form of knowledge” with which the actions will be measured, so that, to a mere application of the data, according to the inserted logic, a vision that contemplates ethics as the addendo of the operation is substituted!
Artificial intelligence closely touches the cognitive computing (cognitive computing), relative to software and hardware, which, simulating the human brain, should improve the decision-making processes; automatic learning (machine learning), which, starting from samples obtained from a set of data combined through algorithms (training data, through which the system learns to recognize models and derive rules, as for deep-learning), creates inductive models and makes predictions from scratch; augmented intelligence, since AI would increase human intelligence rather than replace it.
Therefore, the concept of close collaboration between the human element, capable of judgement, and the technological element must be strengthened; in cases of gender differences, or ethnic and social origin, which can produce different bias of evaluation (defined as “a form of distortion of evaluation caused by prejudice”), it will be necessary that the data, on which Artificial Intelligence rests, do not induce the latter to perform evaluations or actions that may vary, depending on the category of population.
Where is the Artificial Intelligence?
The S.A.R.I., acronym which stands for Automatic System of Image Recognition, is a system at the disposal of the State Police to counteract the criminal activity; exploiting the A.F.I.S. system, Automated Fingerprint Identification System (or Automated Fingerprint Identification System) which collects the fingerprints, personal data, photographs and biometric notations of the subjects under investigation, law enforcement agencies can count on an identification system with a database of more than 10 million data; in this way those who stain a crime can be identified more quickly and efficiently.
In the United States, another system is used, the C.O.M.P.A.S. (Correctional Offender Management Profiling for Alternative Sanctions), which is an algorithm used by judges to calculate the probability of recidivism within two years of a crime: the American Criminal Courts have had an appeal filed, following a conviction of recidivism, for a black prisoner, because the system favoured whites over African-American ethnic groups.
From an ethical point of view, Artificial Intelligence poses problems of transparency and openness, since, often, it is not possible to determine either the data, on which it bases its functioning, or the architecture of its algorithms, since they are often covered by industrial secrecy. Artificial Intelligence has, moreover, prohibitive costs for most companies; it is currently available only to multinationals and governments with high economic resources.
As early as 2017, the South Korean Government ordered that government ministries and agencies should present new legal standards in order to define the legal status, responsibilities and ethical profiles of industries involved in artificial intelligence production and studies.
The primary objective is to start thinking about how to use AI, but also, and above all, how to regulate it, so that it does not govern us, but rather helps us in our daily lives.
Professor David Weinberger, an American thinker and philosopher, writes: “We should all be concerned that artificial intelligence (AI) systems based on machine learning, to which we entrust decisions — from whom to make them — may make serious mistakes, harming individuals and whole categories of people. Sometimes the mistakes are relatively harmless.[…]It is devastating when, in the US, an AI recommends that black men should be kept in prison longer than white men simply because of their race.
What in an old film of 1983, War Games — War Games, seemed an indefinite and distant future, is a concrete reality of our days, with which everyone, sooner or later, will have to measure themselves; the ability of a computer, to learn from their mistakes and to correct themselves, although it represents the dawn of a new era, should not be underestimated, should be understood but “polite”, not to be subjugated by it.
The question about Intellectual Property and copyright
This long and dutiful premise anticipates and raises a nodal point, which we must quickly think about, in a current apparent normative void, that is the applicability of copyright rules, if, as in recent “robotic” creations of pop music albums, or perfect reproductions of works by famous painters, the author in question is not a human being but a machine, and not a machine whose software can be traced back to a human figure, but a new sentient being, who through neural developments of autonomous learning, gives birth to a creation exclusively his own.
Microsoft, in collaboration with the University of Deift and the Rembrandt House Museum in Amsterdam, with a 3D printer, has recreated, with the same depths and color stratification of an original, a new painting, which simulates the same technique and creative line of the original author, called The Next Rembrandt, opening a chasm of questions, not only legal, but exquisitely ethical, on the value of “creative genius”.
To date, any software, even imitative of the human intellect, lacks a legal subjectivity, so who should we recognize, in the field of intellectual property, the ownership of the relevant rights of economic use or even the responsibility for possible plagiarism or counterfeiting of the works of others?
Intellectual property refers to the creations of the mind, such as inventions, literary and artistic works, drawings, and symbols, names and images used in trade. The rights give the owner of the work exclusive rights of exploitation and benefit from its creation.
Perhaps we are at the dawn of a new legal entity, the “robot”, to which, however, it would not be sufficient to apply, in full, Article 6 of the copyright law (L.633/1941) which provides: “The original title of the acquisition of copyright is the creation of the work, as a particular expression of intellectual work. It would be out of any possible recognition, at least moral, of authorship of the work, the possibility of economic exploitation resulting from it!”
At the recent conferences, in which the question was raised, the most compromising solution seemed to be that of attributing such economic exploitation to the person who created the software, being also responsible in case of possible harmful actions against third parties.
In our legislation such recognition is more difficult, since Art. 2580 c.c. (“Copyright belongs to the author and his successors in title within the limits and for the effects established by special laws”) would seem to exclude ownership of a software, distinct from its creator or owner.
In contrast, the UK Copyright, Designs and Patent Act (CDPA) 1988 states that the author of the work produced by software (excluding moral rights) is the one who has made the necessary arrangements for its creation, provided that the work is “original and meets the author’s own “intellectual creation” requirement, reflecting the author’s personality and creative choices; nevertheless, the United Kingdom legislation always specifies that the reproduction of a copyrighted work is prohibited by any electronic means, including even those works which do not represent the final work itself and therefore reproducing, in full process and development IA, a copyrighted work, would in any case entail an infringement.
Unfortunately, the problems do not end there: on October 23, 2018, the 40th International Conference of Privacy and Personal Data Protection Guarantors (“the Conference”) was held in Brussels, during which the Declaration on Ethics and Data Protection in Artificial Intelligence was approved. It was noted that the need, on the part of the development of AI, to process a series of data, mostly personal, creates a serious privacy problem.
Therefore, a set of key points/principles to be respected in the development of IA has been developed:
- Equity (Development consistent with the original intentions and far from possible deviations)
- Attention to the effects and consequences of use
- Ethics from the very beginning of development
- Accountability (Responsibility and subsequent accountability to the human being)
- Impartiality (reducing prejudice and discrimination resulting from the use of data)
Artificial intelligence must also take on this task, to guide an ethical as well as technological development, precisely because algorithms, neural networks, also absorb prejudice (an experiment on facial recognition showed to be 99% accurate on white male males, but only 34% on dark-skinned women). Neural networks lack two exquisitely human components, understanding and intuition.
Since the end of 2017, several “round tables” have been held, the topic of AI has been discussed in many forums, institutional and not, national and international, and one adjective has always been reiterated, with vigor and strength, “trustworthy”, that is, reliable, worthy of trust, and to satisfy this primary need, IA must be legal, ethical and solid and put man at the centre of his development, placing his freedom and well-being as a starting point and as a goal, so that the famous phrase of the great Stephen Hawking does not turn out to be prophetic: «Artificial Intelligence will be man’s most important achievement, a pity it may be his last.»
All Rights Reserved
Raffaella Aghemo, Lawyer