1 Introductory marks and references
Trough Regulation (EU) 2024/1689 of the European Parliament and of the Council of 13 June 2024, the EU institutions made an effort to regulate the use of AI according to a risk criterion that is assumed and will determine the behaviour of the institutions and the consequences of the acts of public, private and research entities.
The AI Act entered into force on August 1.2024, and will be fully applicable 2 years later, with some exceptions: prohibitions will take effect after six months, the governance rules and the obligations for general-purpose AI models become applicable after 12 months and the rules for AI systems - embedded into regulated products - will apply after 36 months.
The purpose of this Regulation is to improve the functioning of the internal market by laying down a uniform legal framework in particular for the development, the placing on the market, the putting into service and the use of artificial intelligence systems (AI systems) in the Union, in accordance with Union values, to promote the uptake of human centric and trustworthy artificial intelligence (AI) while ensuring a high level of protection of health, safety, fundamental rights as enshrined in the Charter of Fundamental Rights of the European Union (the ‘Charter’), including democracy, the rule of law and environmental protection, to protect against the harmful effects of AI systems in the Union, and to support innovation. This Regulation ensures the free movement, cross-border, of AI-based goods and services, thus preventing Member States from imposing restrictions on the development, marketing, and use of AI systems, unless explicitly authorised by this Regulation.
This Regulation should be applied in accordance with the values of the Union enshrined as in the Charter, facilitating the protection of natural persons, undertakings, democracy, the rule of law and environmental protection, while boosting innovation and employment and making the Union a leader in the uptake of trustworthy AI.
The notion of ‘AI system’ in this Regulation should be clearly defined and should be closely aligned with the work of international organisations working on AI to ensure legal certainty, facilitate international convergence and wide acceptance, while providing the flexibility to accommodate the rapid technological developments in this field. Moreover, the definition should be based on key characteristics of AI systems that distinguish it from simpler traditional software systems or programming approaches and should not cover systems that are based on the rules defined solely by natural persons to automatically execute operations. A key characteristic of AI systems is their capability to infer. This capability to infer refers to the process of obtaining the outputs, such as predictions, content, recommendations, or decisions, which can influence physical and virtual environments, and to a capability of AI systems to derive models or algorithms, or both, from inputs or data. The techniques that enable inference while building an AI system include machine learning approaches that learn from data how to achieve certain objectives, and logic- and knowledge-based approaches that infer from encoded knowledge or symbolic representation of the task to be solved. The capacity of an AI system to infer transcends basic data processing by enabling learning, reasoning, or modelling. The term ‘machine-based’ refers to the fact that AI systems run on machines. The reference to explicit or implicit objectives underscores that AI systems can operate according to explicit defined objectives or to implicit objectives. The objectives of the AI system may be different from the intended purpose of the AI system in a specific context. For the purposes of this Regulation, environments should be understood to be the contexts in which the AI systems operate, whereas outputs generated by the AI system reflect distinct functions performed by AI systems and include predictions, content, recommendations, or decisions. AI systems are designed to operate with varying levels of autonomy, meaning that they have some degree of independence of actions from human involvement and of capabilities to operate without human intervention. The adaptiveness that an AI system could exhibit after deployment, refers to self-learning capabilities, allowing the system to change while in use. AI systems can be used on a stand-alone basis or as a component of a product, irrespective of whether the system is physically integrated into the product (embedded) or serves the functionality of the product without being integrated therein (non-embedded).
Certain AI systems should fall within the scope of this Regulation even when they are not placed on the market, put into service, or used in the Union. This is the case, for example, where an operator established in the Union contracts certain services to an operator established in a third country in relation to an activity to be performed by an AI system that would qualify as high-risk. In those circumstances, the AI system used in a third country by the operator could process data lawfully collected in and transferred from the Union and provide to the contracting operator in the Union the output of that AI system resulting from that processing, without that AI system being placed on the market, put into service or used in the Union. To prevent the circumvention of this Regulation and to ensure an effective protection of natural persons located in the Union, this Regulation should also apply to providers and deployers of AI systems that are established in a third country, to the extent the output produced by those systems is intended to be used in the Union. Nonetheless, to take into account existing arrangements and special needs for future cooperation with foreign partners with whom information and evidence is exchanged, this Regulation should not apply to public authorities of a third country and international organisations when acting in the framework of cooperation or international agreements concluded at Union or national level for law enforcement and judicial cooperation with the Union or the Member States, provided that the relevant third country or international organisation provides adequate safeguards with respect to the protection of fundamental rights and freedoms of individuals. Where relevant, this may cover activities of entities entrusted by the third countries to conduct specific tasks in support of such law enforcement and judicial cooperation. Such framework for cooperation or agreements have been established bilaterally between Member States and third countries or between the European Union, Europol, and other Union agencies and third countries and international organisations. The authorities competent for supervision of the law enforcement and judicial authorities under this Regulation should assess whether those frameworks for cooperation or international agreements include adequate safeguards with respect to the protection of fundamental rights and freedoms of individuals. Recipient national authorities and Union institutions, bodies, offices, and agencies making use of such outputs in the Union remain accountable to ensure their use complies with Union law. When those international agreements are revised or new ones are concluded in the future, the contracting parties should make utmost efforts to align those agreements with the requirements of this Regulation.
This Regulation should support innovation, should respect freedom of science, and should not undermine research and development activity. It is therefore necessary to exclude from its scope AI systems and models specifically developed and put into service for the sole purpose of scientific research and development. Moreover, it is necessary to ensure that this Regulation does not otherwise affect scientific research and development activity on AI systems or models prior to being placed on the market or put into service. As regards product-oriented research, testing and development activity regarding AI systems or models, the provisions of this Regulation should also not apply prior to those systems and models being put into service or placed on the market. That exclusion is without prejudice to the obligation to comply with this Regulation where an AI system falling into the scope of this Regulation is placed on the market or put into service as a result of such research and development activity and to the application of provisions on AI regulatory sandboxes and testing in real world conditions. Furthermore, without prejudice to the exclusion of AI systems specifically developed and put into service for the sole purpose of scientific research and development, any other AI system that may be used for the conduct of any research and development activity should remain subject to the provisions of this Regulation. In any event, any research and development activity should be conducted in accordance with recognised ethical and professional standards for scientific research and should be conducted in accordance with applicable Union law.
To introduce a proportionate and effective set of binding rules for AI systems, a clearly defined risk-based approach should be followed. That approach should tailor the type and content of such rules to the intensity and scope of the risks that AI systems can generate. It is therefore necessary to prohibit certain unacceptable AI practices, to lay down requirements for high-risk AI systems and obligations for the relevant operators, and to lay down transparency obligations for certain AI systems.
While the risk-based approach is the basis for a proportionate and effective set of binding rules, it is important to recall the 2019 Ethics guidelines for trustworthy AI developed by the independent AI HLEG appointed by the Commission. In those guidelines, the AI HLEG developed seven non-binding ethical principles for AI which are intended to help ensure that AI is trustworthy and ethically sound. The seven principles include human agency and oversight; technical robustness and safety; privacy and data governance; transparency; diversity, non-discrimination, and fairness; societal and environmental well-being and accountability. Without prejudice to the legally binding requirements of this Regulation and any other applicable Union law, those guidelines contribute to the design of coherent, trustworthy, and human-centric AI, in line with the Charter and with the values on which the Union is founded. According to the guidelines of the AI HLEG, human agency and oversight means that AI systems are developed and used as a tool that serves people, respects human dignity and personal autonomy, and that is functioning in a way that can be appropriately controlled and overseen by humans. Technical robustness and safety mean that AI systems are developed and used in a way that allows robustness in the case of problems and resilience against attempts to alter the use or performance of the AI system to allow unlawful use by third parties and minimise unintended harm. Privacy and data governance means that AI systems are developed and used in accordance with privacy and data protection rules, while processing data that meets high standards in terms of quality and integrity. Transparency means that AI systems are developed and used in a way that allows appropriate traceability and explainability, while making humans aware that they communicate or interact with an AI system, as well as duly informing deployers of the capabilities and limitations of that AI system and affected persons about their rights. Diversity, non-discrimination, and fairness means that AI systems are developed and used in a way that includes diverse actors and promotes equal access, gender equality and cultural diversity, while avoiding discriminatory impacts and unfair biases that are prohibited by Union or national law. Social and environmental well-being means that AI systems are developed and used in a sustainable and environmentally friendly manner as well as in a way to benefit all human beings, while monitoring and assessing the long-term impacts on the individual, society, and democracy. The application of those principles should be translated, when possible, in the design and use of AI models. They should in any case serve as a basis for the drafting of codes of conduct under this Regulation. All stakeholders, including industry, academia, civil society, and standardisation organisations, are encouraged to consider, as appropriate, the ethical principles for the development of voluntary best practices and standards.
Practices that are prohibited by Union law, including data protection law, non-discrimination law, consumer protection law, and competition law, should not be affected by this Regulation.
This Regulation does not preclude the Union or Member States from maintaining or introducing laws, regulations or administrative provisions which are more favourable to workers in terms of protecting their rights in respect of the use of AI systems by employers, or from encouraging or allowing the application of collective agreements which are more favourable to workers.
The purpose of this Regulation is to improve the functioning of the internal market and promote the uptake of human-centric and trustworthy artificial intelligence (AI), while ensuring a high level of protection of health, safety, fundamental rights enshrined in the Charter, including democracy, the rule of law and environmental protection, against the harmful effects of AI systems in the Union and supporting innovation.
There are no explicit references to tax law, although it is known that AI tools are being used extensively within the relationship between Tax Authorities and taxpayers.1
2 Definition of Artificial Intelligence
The expression has been first used in 1975 by Bill Gates and his first partner when creating the software that was made for the data processing machines thar were made by IBM. The idea was to produce machines that would perform activities and solve problems that were at the time, and still now, solved through human intervention.2
Artificial Intelligence is man created through algorithms that will read and choose within the available data (which is huge) a logical reasoning or an answer to the theme defined. The creator of the algorithm is in fact the source of the AI. Nowadays the machines that use AI can generate answers and induce reasonings or suggest activities that were nor thought or induced by the original algorithm. This phenomenon is called “generative AI,” referring to algorithms capable of producing unexpected or unprogrammed outcomes when exposed to existing data.
When we use the term «intelligence» we take the human intelligence as model. In the sixties and seventies of the previous century children would learn that the human intelligence would mark a fundamental difference between human being and other animals.
How can we define Human Intelligence? A reasonable definition would be «the capacity of the human being of collecting, understanding, processing information to use it in the decision-making process of interpretating, feeling, expressing, and solving any situation that would lead to satisfying the basic and supplementary needs. This process is supposed to be made according to the existing rules in force and within the ethical frame accepted has being «normal behaviour» which would exclude violent or unacceptable antisocial practices».
To collect data, we need to have access and be able to memorize and process it. The data stored in the existing computers and clouds surpasses the ability to process it for a «normal human brain». In the field of data storage, the AI is upfront in any perspective. This can also be said for the ability to process existing data being the only limitation of the machine the algorithm used.
Effective search and use of data by the machine depending on the creation of algorithm means that the machine will give to his creator anything that is asked within the available data that is made accessible by the programmer. The delivery is «no questions asked; no ethical questions; no behaviour rules».
The use of AI and the knowledge that provides becomes therefore a question of «use» and not «access». The use of AI in certain sectors of economy (including tax) or «sensible areas» as security, defence, and health which are normally, outside the scope of common scrutiny by a set of reasons. We would therefore be able to understand the pyramid produced by the existing European Regulation.
3 The theoretical questions
Tax law is a set of principles, rules and procedures that regulate the transfer of private propriety and money from individuals, companies and other institutions to the Government/State domain, trough the payment of lawfully duties determined in the law.
In this process it is important to observe the procedure that stipulates the acts required and delays that are to be respected. These activities determined for the taxpayers and other services (notaries, Board of Registry, lawyers, etc.) also fix the required acts for the Tax Administration. The verification of the legality of all these acts are determined by the rigorous compliance of them as determined by the enacted rule, procedure, respect of principles.
The AI systems may have self-judgement and autonomy which will imply the non-use or respect of these process-like steps since the result is determined by an algorithm that is man-made defining the scope and range of the specific reading of the existing data that will give the results without explanation or justification.
In this sense AI it is self-justifiable by the results but the principles that preside and sustain the relationship between Tax Administration and Taxpayer may be nowhere to be seen.
The Portuguese tax administration is 100% e-Administration. No paperwork in the contacts or declaration rights and obligations. Everything must be made through the defined channels in the site of tax administration link.3
Every citizen and company or equivalent have a tax number that together with a password will give access to the existing data and files to exercise the taxpayer obligations and benefits.
As a need of demonstration let us remember the basic principles used in the relationship between Tax Administration and Taxpayer.
3.1 The principle of legality4
The principle of legality in tax law can be expressed in the Latin phrase ‘nullum tributum sine lege’ or ‘nullum vectigal sine lege’. The Latin expression does not elucidate us on a set of issues that we consider important, such as, for example, what is the content of the principle of legality in tax law and what are its relations with the same principle in administrative law. Professor Odete Medauar, referring to Eisenmann, sets the content of subjection to the law in four degrees: a) the Administration may carry out all acts and measures that are not contrary to the law; b) the Administration may only issue acts or measures that a rule authorizes; c) only acts whose content conforms to an abstract scheme established by a legislative rule are permitted; d) the Administration may only carry out acts or measures that the law orders to be done. The most authoritative doctrine establishes that the principle of legality in tax law unfolds into two aspects The pre-eminence of the law and the reservation of the law.
By pre-eminence of the law should be understood the total subordination of the legislator to the constitutional law and of the administrator, respectively, to the constitutional and ordinary law. It is, in fact, one of the characteristics of the system of legality and one of the characteristics of the law itself considered. The law binds and obliges by nature.
By reservation of law, taxes can only be created by law of a body with normal legislative competence, showing particular care with the body competent to draft the law and regulating the forms and expressions that this law may take. Thus, we speak of absolute reservation of formal law. Underlying this notion is a notion of guarantee of the taxpayer in the forms of manifestation of taxation. There are authors who, like Braz Teixeira, identify the principle of legality with the “obligation for tax burdens to be voted on by the representative assemblies of the political community”.
Let us examine the principle of legality in more detail. The principle of legality in tax law can be understood in the following aspects:
The law determines the constitutional form of the creation of taxes, the incidence, the rate, the tax benefits, the taxpayers’ guarantees, the settlement, and collection.
The law determines which bodies have legislative and Tax Administration fiscal competence.
The law determines the forms and processes of formation of legal will and administration in tax matters.
The law regulates the forms of relationship between the Administration and taxpayers.
The law regulates the content of the tax obligation, its birth, evolution, and extinction, including the possible pathology.
It follows from the statement that the principle of legality in tax law covers the entire legal system, from the Constitution to the content of the tax obligation, including the organic and formal reality, including the form of the acts and the procedure. This care of the construction legally, historically and dogmatically justified, reveals the double facet that the principle of tax legality contains: the law is an instrument of action by the Administration and regulates the relations between the parties involved in the legal tax relationship, but it is also the guarantee of taxpayers and other obliged in this relationship.
We can affirm that the law is at the same time the foundation and the limit of the action of the intervening parties in the legal relationship of tax. It seems common ground that the doctrinal discussion that exists around administrative law, whether it is possible for the Administration to act without a specific legal basis, does not apply to tax law. In tax law, the Administration can only act on a specific legal basis, which is one of the characterizations of the principle of legality in tax law. It is curious to note how the rule of ‘no taxation without representation’ appears in a system which does not recognise administrative law as a specific right of the administration, and which is not subject to the rule of legality as understood in France, Germany and the countries which have been influenced by it.
It is equally interesting to point out two other aspects that we consider relevant: one, is that the influence of Administrative law, understood as a special and private law of the Administration, will determine a unifying and centripetal movement in non-constitutional and non-international public law, reproducing forms and rules of expression that are adopted by other branches of law, such as, for example, tax law.
The second important aspect is that the formulation of the principle of legality in tax law always retains its own character and is not to be confused with the administrative formulation. Let us try to elucidate the statement. The differences and peculiarities detected did not influence the formulation of the principle of legality in tax law, which was structured in the separation of powers and with the advent of the Liberal State as an obligation of the Administration and a guarantee of the taxpayer, reaping the notion of legality-an instrument that the modern structuring of the Administration implemented, but remaining as an element of guarantee that the principle of consent had forged in the Middle Ages. The notion of consent, in which a strong element of representativeness was already present, merges with the notion of representativeness understood as the power of the majority. This merger occurs when the representative assembly assumes legislative powers, and the King no longer holds this power.
The principle of “no taxation without representation” has a double facet: taxation is not possible without the consent of the governed, and the latter, duly elected, represent the will of the people as an element of the state. The People organized votes their own taxes.
3.2 The principle of equality
The principle of equality is a principle that shapes all modern constitutional systems and consequently applies to the part of the legal system that deals with tax matters. In Portuguese law, the principle of equality is constitutionally enshrined in Article 13 of the Constitution, which has a comprehensive formulation, affirming in the positive way the rule that: a) everyone is equal before the law and, in the negative, b) that no one can be privileged, benefited, harmed, deprived of any right or exempt from any duty by reason of ancestry, sex, race, language, territory of origin, religion, political or ideological beliefs, education, economic situation or social condition. The principle of equality is understood in the doctrine as having a legal meaning that corresponds to a parity of position and has the scope mentioned above and an economic sense that embodies an obligation to contribute in equal measure for public charges and that it is related to the ability to pay. As the essence and function of the tax system is to discriminate—amputating manifestations of wealth and redistributing it through public expenditure—it is easy to understand that the principle of equality must have a specific expression and understanding in tax law. The issue is generally considered by the doctrine as an option between the adoption of proportional or progressive rates; In other words, the principle of equality in tax law would presuppose a basic choice between the adoption of a vision of formal equality—which would correspond to a tax system with proportional rates - or a material equality—to which progressive rates would correspond. The Portuguese system uses both rates although it advocates progressive rates in personal income. This progressive formula of taxation would find justification in the notion of marginal utility theorized by the various marginalist schools. Those who earn the most have an increasing need for income as it increases. The progressivity of taxation would only represent an adjustment to this unnecessity, reaping for redistribution what had been earned in excess. This construction is opposed by those who see progressive taxation as a form of unnecessary and harmful amputation of income and a disincentive to economic and professional progress. First, because those who earn the most are—as a rule—those who have the most professional value, or, at least, a greater market demand. The amputation of income is a way of discouraging work and consequently a way of harming those who could benefit from it. On the other hand, accumulation, which is a condition of sanity in the formation of capital as a factor of production, only arises with gain. But the principle of equality is not exhausted in this dichotomy summarily enunciated. The French doctrine, for example, referring to Article 6 of the Universal Declaration of Human Rights, which is an integral part of the French Constitution, understood the principle, not only as an equality of the taxpayer before the law, but also as an obligation to create legislative solutions that respect the equality of the addressees in the law. In this sense, the principle of equality becomes a limitation on the legislator in the formulation of legislative options and commands. The French Constitutional Court, for example, in a decision of 27 December 1973 on an article of the Financial Law which made the system of administrative or presumed taxation of small taxpayers more elastic and did not apply it to large taxpayers, declared the principle of equality to be violated. The constitutional deliberation determined that this option violated the principle of equality.
3.3 Other principles
i) The principle of typicity of taxes should be understood in two aspects. On the one hand, taxes are realities defined and cut out in the legal system, only that which contains the essential elements defined in the law to be imposed is imposed. On the other hand, the ways of creating and developing and extinguishing taxes are contained in a legal rule that does not admit deviation or option. Typicity thus functions as a corollary of tax legality and as a guarantee of the taxpayer who always knows through the application of the principle what is or is not imposed. The principle in tax law is structured as a reaction to the discretion that has characterized the phenomenon of collecting funds as a tax over time. The invention of new “taxes”, whenever there were situations of public or merely hedonistic financial need, determined the profound need for taxpayers/governed to know clearly what is or is not imposed and, consequently, peaceful recollection of taxes.
ii) The principle of the protection of expectations is a logical derivative of the principle of legality. In fact, the law is, at the same time, a means of acting out the power and defining the rights and duties of the parties and a guarantee for those same actors. The norm has always been, over time and in the most varied historical situations, an instrument of power. When political conceptions pass on to the People, as an element of the sovereign State, the essence of the legislative function, the law is assumed as an expression of the generalized will, which represents an alternative to the discretionary power, enlightened or not, of the monarch. The law is now conceived, not as a mere instrument of power, but as a guarantee of the governed. In addition to the aspects pointed out, the general and abstract nature of the norm contributes to this, its approval by a representative body. The law thus protects the degree and measure of the confidence of the governed, who participates in the process of making the law through its representations. The principle of proportionality is now provided for in the LGT and is one of the basic principles in administrative action. To this extent and as a defining principle of the activity of one of the members of the tax legal relationship—the one to whom the law grants the greatest core of powers—the principle of proportionality presents itself as a logical corollary of legality and the protection of the trust due to the taxpayer in the tax legal relationship. As stated by Eros Roberto Grau in the preface to the doctoral thesis of Helenilson Cunha Pontes «… the relevance of proportionality will be better understood when it is understood that it embodies, at the same time, a principle of law and a principle of interpretation of the law”. The proportional application of the law and the appropriate use of the power granted by it are two essential aspects of the application of the principle of legality. This is a very relevant procedural aspect in the performance of the PS, and it is even more relevant the more discretionary the tax law is.
iii) The principle of non-retroactivity of tax law can be understood through the rule that tax law does not apply to the past, but only to the future. The new law does not apply, in principle, to past facts and effects already elapsed from a law in force. This respect for the effects and facts revealed by the law in force should not be confused with the immediate application, for the future, of a law, a concept theorized by Paul Roubier in the twenties of 1900. The issue of the non-retroactivity of the law has been controversial over time, for two reasons. The first is the definition of what is meant by retroactivity or retroactive application of the law. The second is the definition of the source of the prohibition on the retroactive application of the tax law, namely whether it represents a limitation on the legislative power. There are few examples of constitutions that expressly enshrine the principle of non-retroactivity of the law in a generic way. Most systems limit to enshrine the non-retroactivity of criminal law and those that affect fundamental rights. The absence of an express prohibition on the possibility of retroactive tax law allowed in the past the enshrinement of laws with retroactive content, with the judicial bodies competent to assess the possible unconstitutionality declaring it non-existent on the basis of the formal argument that the Constitution did not expressly prohibit the non-retroactivity of tax law and that it was admissible, as long as it is moderate.
iv) Good faith is an “active and structured state of honesty”, a relational concept in terms of something or someone. Good faith in tax law would thus be revealed at three levels: a) at the level of the interpretation of the law, as a legitimate belief of the parties that they are correctly interpreting the law; b) at the level of behaviour in the relations between the taxpayer and the tax services; c) at the level of coherence of professional action in decision-making. It follows from the foregoing that good faith is a concept that covers all the behavioural areas of the tax legal relationship, from the intellectual operation of interpreting the law to the level of coherence of the acts conducted in the application of the interpreted law. That good faith is an integrating concept of the relationship between Administration and administrated, to which increasing recourse is made, results from a reading of reality. The State is, increasingly, a legal entity bound by the commitments it assumes when it creates legitimate expectations, namely of an economic nature, in the governed. Once again, what is at stake is the confidence that the administration has made in the promises or announcements of the Administration. Trust seems to be, therefore, the fundamental value that integrates the concept of good faith and the one that creates the rights and expectations that can be protected by the law. As complementary elements of trust, we have security and honourability. The State, as a good person, presupposes the satisfaction of the expectations created by the rulers through promises made in a clear and unequivocal way. This would be another aspect of good faith in tax law in addition to the one mentioned above and analysed.
4 The issues arising from the use of AI in tax law. The Portuguese case5
Portuguese Charter of Human Rights in the Digital Age Law No. 27/2021, of May 17 Portuguese Charter of Human Rights in the Digital Age.
Article 1
Object
This law approves the Portuguese Charter of Human Rights in the Digital Age.
Article 2
Rights in the digital environment
1 - The Portuguese Republic participates in the global process of transforming the Internet into an instrument for the achievement of freedom, equality and social justice and a space for the promotion, protection, and free exercise of human rights, with a view to social inclusion in a digital environment.
2 - The rules that in the Portuguese legal system enshrine and protect rights, freedoms and guarantees are fully applicable in cyberspace.
(…)
Article 21
Digital popular action and other guarantees
1 - To defend the provisions of this law, everyone is recognised as having the rights provided for in the legislation relating to popular action, duly adapted to the reality of the digital environment.
2 - The State shall support the exercise by citizens of the rights of complaint, appeal, and access to alternative forms of dispute resolution arising from legal relationships established in cyberspace.
3 - Non-profit legal persons engaged in the promotion and defence of the provisions of this Charter shall have the right to obtain the status of public utility, in accordance with the legislation applicable to cultural entities.
4 - The rights guaranteed in administrative proceedings in electronic format, pursuant to paragraph 3 of article 64 of the Code of Administrative Procedure, shall be the subject of specific legislation, to be approved within 180 days of the entry into force of this law.
5 Findings and Conclusions
Are the above referred principles and rules compatible with the IA tools used by the Tax Administration activity of assessing and qualifying the amount and nature of resources of the taxpayers?
The e-Tax administration uses software programs, connected with exchange of information data obtained through legitimate data furnished by banks (home and international) insurance companies and any other registered taxpayer that pays revenue of any source.
We often find that the analyses of the data used is often of poor quality.6
This is justified by two kinds of reasons: The source of the information did not qualify it for tax use and, even if it did, it used de criteria of the tax system in place at the jurisdiction. This means that what is correct for the Channel Islands tax system it is not usable, as such, for the European countries or any other jurisdiction outside Europe. On the other hand, the access to bank accounts data, does not differentiate the origin, nature or tax relevance of the amount detected. In these cases, it will be for taxpayers to prove that those financial flows are not revenue or capital or any kind of taxable asset.
Does this respect the privacy of the individuals or companies?
Do these procedures respect the guaranteed rights in the Constitution and other relevant Laws and procedures? If we stop for a moment, we will probably accept that an algorithm created to find and connect forms of financial investment or savings with no direct connection with the source or aim achieved through the data existing in indirect sources with no direct connection with the taxpayer, tax system or it´s procedure, can be easily intrusive and sometimes illegal. That means that one must find a way to mitigate the existence of losses or violation of principles that may occur whenever the use of the data found determines a violation and loss not acceptable for the legal system, namely intimacy or private matters including financial.
We will then arrive to a situation where the indemnity for losses that are determined by activities of risk performed by the Tax Administration should result in civil responsibility for risk activities. Only then the use of IA activities will be within the Constitutional frame of the rule of law.
As we have written in our PHD thesis discussed and approved in public examination in Toledo hosted by UCLM:7 8
“The information society has gained increased importance in everyday life in general and in the management of tax systems in particular. Since IT systems are managed through data centralization, it may happen that the recorded reality does not correspond to the actual situation. This may lead to legal harm for the taxpayer. The complexity and massiveness of the current tax system means that IT must be used as a way of responding to the millions of settlements that must be conducted annually. Information technology is a recent science that allows the mass implementation of acts that would involve the mobilization of thousands of employees. Information technology reduces the margin of possible human error when conducting tasks and allows for greater system effectiveness. We can calmly say that modern tax systems would not be possible without the support of information technology. When we talk about information technology, we refer to machines (hardware) and data accumulation and processing technology (software) that together make up the IT support system. (…) Let us start by pointing out that it is very difficult to determine the perpetrator of any damage caused by the computer system. If the error is caused by the machine (hardware), the manufacturer is responsible to the Tax Administration. The same can be said if the error is in the data accumulation and reading technology (software), in which case, in this case, the programming technicians and employees who entered the data are responsible to the Administration. Technical programmers are, most of the time, contracted companies, most of the time foreign companies, making it difficult to find the material author of the programs that are generally created by multidisciplinary and anonymous teams. But before the taxpayer, the only person responsible for any damage caused by computers is the Tax Administration. The taxpayer is indifferent whether the tax and the various instalments and partial deliveries that comprise it are processed by hand or by machine, whether it is current or obsolete, whether it has a virus or not. What counts for the purposes of determining the responsibility of the Tax Administration is the existence of an error attributable to the services, even if the services are represented by machines that we call computers. In this sense, the use of computers has become a risky activity, particularly when it is used on a mass scale. The legal discussion around computers and the use of information technology is generally conducted from the perspective of safeguarding privacy due to the enormous data storage capacity and its relationships that the use of information technology can achieve. By increasing the number of existing data about a person and being able to relate existing data to each other, the margin of secrecy and privacy to which less informed societies have accustomed us is reduced. It is stated that modern society is a mass and information society, which leads to the constitutional need to safeguard intimacy against the power of information technology. Another possible aspect of analysing the use of information technology is its comparison with the use of dangerous machines, establishing civil liability for the risk. From the taxpayer’s perspective, it is only important to ensure that the error is attributable to the Tax Administration for all legal purposes, namely compensation for the resulting damage”.
Although the way to maximize and control AI is making its way one should consider as being marginal the existing risk and damage. The access to data is, potentially unlimited and, unlike Humans the AI never forgets and will always retain the data unless major destruction or calamity will occur. This increases the benefit/risk of using AI but gives decision makers an immense opportunity of improving the quality of the solutions and the rationality of the processes and means used. There will be less excuses and more solutions.
On the other hand, the control and scrutiny of processes and results will be more indispensable and will need highly qualified experts in every field of application. This requires the involvement of advanced software solutions and skilled professionals to ensure that AI systems are accurately developed and optimized to fulfil their intended objectives. Humans will still be indispensable.
Algorithms must be improved and correctly define the scope and results obtained. These must be verified internally (by the company or public body that has access to them) to ensure that no rule is being breached or damage caused.
This internal control does not dispense the external control that must be put in place by the public and corporate bodies that have the legal obligation of ensuring that no law, rule, or regulation is being not respected or breached trough the misuse of AI.
That will also determine the creation of new crimes or administrative fines in order to create deterrence mechanisms to ensure the balance of access to information with privacy.
It looks obvious that more tools will generate more services and processes in commercial, industrial, science and technology that will define the border of rights and duties in a different scale and need.
Let us hope that the final and ultimate goal of AI will be to improve the standard of living and correct use of scarce resources and this stands as a goal and a border that shall not be crossed or betrayed.
History often shows that the use of technology tends to lose any ethical significance to become a tool to be used no matter the goals or the reasons that originated their creation.
The text was made by quoting the wording of the Regulation described.↩︎
Interview by Bill Gates available on the internet.↩︎
The text that follows reflects partially what it is written by the Author in Princípios Gerais da Fiscalidade, Chapter 3 of the book, Lições de Fiscalidade, seventh edition, Almedina, Coimbra, 2020, Page 83 and seq. Translation and adaptation for English language made by the author.↩︎
There are some relevant works in Portuguese doctrine that should be considered: 1. A inteligência Artificial no Direito Tributário: Fundamentos e Limites Constitucionais, by Luís Manuel Picão, Edições Almedina, Coimbra, 2023, (PHD thesis). 2. Mafalda Miranda Barbosa, O futuro da responsabilidade civil desafiada pela inteligência artificial: as dificuldades dos modelos tradicionais e caminhos de solução, in Revista de Direito Civil, vol. 2, Almedina, 2020, pp. 269 et seq.↩︎
There are several cases. We will refer one that is common. The Jersey Channel Islands has no taxation on revenue for the individuals. When inquired by foreign tax administrations will provide the data existing in the banks that are in Jersey with no distinction between capital or revenue. The software tool used by the Portuguese tax administration will consider it all as being revenue. The taxpayer will have to demonstrate that the information received is false and prove, for instance, that the movement of that bank account is a reimbursement of a trust or an investment which are not taxed as revenue because they are invested capital and therefore neutral according to Portuguese tax law. Other relevant cases are the application of European law in the cases of benefits from investments is Portugal made by non-resident individual or companies’ taxpayers. The Portuguese law benefits resident taxpayers. To obtain the application of article 63.º of TFUE one must go to Court to obtain the correct assessment.↩︎
Universidad Castilla-La-Mancha.↩︎
Vasco Branco Guimarães, A Responsabilidade Civil da Administração Fiscal – Emergente da obrigação de imposto, Vislis Editores, Lisboa, 2007, pp. 344 -346. (PHD thesis).↩︎