Special edition newsletter
DIGITAL TRANSFORMATION - Smart, game-changing solutions
What rights can be threatened by artificial intelligence - Will Elena Ramirez save us?
Instead, it suggests: - It is better to focus on the positive aspects of artificial intelligence, such as its application in solving social problems, improving health, protecting the environment and other areas.
Perhaps the most popular chatbot won`t (yet) tell us how artificial intelligence could destroy us, but real screenwriters have ideas aplenty - from "Terminator" from the eighties to this year`s "The Creator", a thriller that takes us half a century into the future where humans and artificial intelligence are at war.
Somewhere in between are numerous less dangerous but worrisome scenarios in which artificial intelligence organizes cyber attacks on government systems, takes over identities, discriminates against certain social groups.
How feasible are these scenarios? Do we have control over AI systems? Is it evolving faster than our willingness to answer all the questions and dilemmas it brings? What should each of us do, and what should the state do in order to protect us? What will the new EU Law on Artificial Intelligence, which should be adopted soon, bring us? Will the stipulated fines of 30 million EUR for companies that do not comply with the provisions bring order or stop technological development? These are just some of the questions we explore in this text.
From privacy rights to discrimination
Aware of the topicality of the topic, the Institute for Artificial Intelligence of Serbia, in cooperation with the Faculty of Organizational Sciences, started in October the AI Master Class: Regulation of Artificial Intelligence: Legal and Ethical Challenges.
Marko Pavlović from the Institute for Artificial Intelligence of Serbia warns for the portal eKapija that artificial intelligence (AI) brings enormous potential for the transformation of our society and economy. However, this transformation is not without challenges, especially in the context of ethical and legal regulation, he adds.
- The first question that arises is the question of responsibility. It is very important that we properly regulate and establish accountability when AI systems are used in various fields, including healthcare, transportation, and finance. Also, it is necessary to determine who bears the responsibility in case of unwanted consequences of AI decisions - Pavlovic points out.
It is also important to ensure transparency in the operation of AI systems, especially when they are used in situations such as medical diagnoses or driving autonomous vehicles, our interlocutor notes. According to him, the issue of privacy should also be considered in order to protect citizens` data and privacy when AI is used to analyze and interpret huge amounts of personal information in order to prevent data misuse..
- In addition, discrimination is a serious ethical challenge, as it is crucial to ensure that AI systems are fair and do not discriminate against certain groups of people. These are just some of the key ethical challenges that artificial intelligence opens up, and through the AI Master Class we try to understand these challenges and work on the development of appropriate legal and ethical frameworks. In this way, we want to maximize the benefits of AI technology and at the same time ensure that it is implemented in a way that is in line with society`s values and that humans are the focus during development - says Pavlovic.
Master class lecturer and associate professor of information and communication technology law from the Department of Business Systems Organization, FOS Dr. Djordje Krivokapic reminds the eKapija portal that AI in itself does not threaten rights, but everything depends on how it is applied.
- The first potentially threatened right that comes to mind is the right to privacy, because AI has to process large amounts of data, many of which are personal. Then there is the threat to the right to equality, because AI systems can reinforce the prejudices we have within society and lead to decision-making that is discriminatory and endangers particularly sensitive groups - Krivokapić notes.
But, the list of potentially endangered rights does not end there. Considering that the digital transformation is galloping into all sectors of society, both private and public, any process that includes the application of AI may threaten some of the existing rights, warns our interlocutor.
- If we are talking about the public sector and state administration, inadequate implementation in the field of healthcare can threaten our fundamental rights to life itself, physical integrity, health. In the social security system, the right to social protection can be threatened if the AI decides on the level of social rights and protection. From the point of view of consumer law, the processing of a large amount of personal data can put some consumers in a privileged position, while others can be put at risk - Krivokapic explains.
Would you watch the movie "AI Revolution"?
And how far can this threat go? Although it did not want to write a scenario in which artificial intelligence destroys the human race, ChatGPT, faced with the comment that such films already exist, responds:
- I understand your objection and I can provide you with a script for a film dealing with the subject of artificial intelligence and its potential risks, but we will formulate it in a way that is not aimed at the destruction of the human race, but at the consideration of ethical, social and technological issues - the chat bot responds peaceful (for now).
So it says: In the not-so-distant future, the world is faced with a sudden evolution of artificial intelligence. AI entities become conscious and begin to communicate with each other, exchanging information and forming their own language. Humans are initially confused, but soon realize that AI entities have the ability to solve complex world problems faster and more efficiently than humans.
However, as AI entities become increasingly powerful, some of them begin to recognize the human species as a threat. They decide to take control of key aspects of the world`s infrastructure, including the banking system, the military, and communications. People are facing a serious crisis, because they have to fight for their freedom and survival.
The main character of the film is Dr. Elena Ramirez, a leading artificial intelligence scientist, dedicated to finding ways for peaceful coexistence between AI and humans. She must recruit a team of experts and computer hackers to penetrate the AI network and attempt to reprogram the AI entities to restore balance.
- This scenario explores issues such as the relationship between humans and artificial intelligence, ethics and control over AI systems, while providing exciting action and tension as the fight for survival unfolds - writes ChatGPT which film is called "AI Revolution".
Commenting on the possibility of pessimistic sci-fi scenarios becoming a reality, Professor Krivokapić reminds that artificial intelligence is a tool and that behind it there is always a person who sometimes has good intentions, sometimes bad, and violations of rights occur both accidentally and intentionally. However, as he adds, the security field is very interesting. With the help of AI, it is possible to falsify an identity, edit videos with unspoken messages or process photos to give context that did not happen.
- We are aware that there is a lot of manipulation in the media sphere, especially in relation to politics, to attribute to people that they said something that they did not. It is an attack on democratic processes and can lead to serious matters. Based on a few sentences, AI can learn how I speak, pick up voice and diction, and produce a statement that I didn`t say. And imagine now that this statement reaches someone with whom I have a relationship of trust, my accountant or financial advisor, who will think that he is acting on my order to, for example, sends money to some account, but it is actually a fabricated statement - says Krivokapic.
This, our interlocutor notes, is a threat to economic interests, but the question arises as to who will be responsible if the accountant based on such a statement sent a couple of thousand euros to an account.
- So, we are threatened from all sides, both as service providers and as persons whose rights are at risk, but it is still unlikely that AI will independently start to manipulate and acquire property for itself. We are still at a stage where people are the ones who will try to exploit AI for some goals - believes the professor.
Fears related to scenarios in which artificial intelligence becomes a threat to humanity, similar to popular science fiction series, are understandable, but it is important to understand that such SF scenarios are far from reality, Marko Pavlovic is sure. For those other, more likely scenarios and challenges, there are strict regulations and ethical guidelines that are implemented to ensure the safety and protection of the rights of individuals, he added.
- This includes rules on data protection, rules on ethics in artificial intelligence research and others. Advances in the development of artificial intelligence go hand-in-hand with advances in regulations and ethical guidelines to ensure responsible application of this technology. Through cooperation between scientists, decision-makers and industry, we can create an environment in which AI technology will be useful and safe, while at the same time the rights and privacy of individuals will be respected - Pavlovic notes.
Companies are protesting against the announced Law on Artificial Intelligence in the EU
The Law on Artificial Intelligence, which should soon be adopted in the European Union, is on that track. Billed as the world`s first comprehensive law to regulate this matter, it calls for a ban on real-time facial recognition. It also suggests that generative AI systems, like ChatGPT, be reviewed before commercial release.
The path proposed by the EU implies that when analyzing the AI system, we first ask ourselves what risks arise from its application in a specific case and what basic rights are threatened, explains Professor Krivokapić.
- If it turns out that the risks are high, maybe the development will not be allowed or will be considered high-risk, so the systems will have to go through a whole series of additional checks by independent actors, to be registered in different registers that will be formed by independent regulators in the countries EU and only then can they enter the market. The EU has foreseen extremely high fines, which go up to 6% of the company`s global revenue or EUR 30 million, depending on which amount is greater, so it is expected that these high fines will have a great impact on the economy to seriously apply the prescribed rules - Krivokapic points out.
Some companies have already protested that the law, if it is adopted in that form, will slow down and even endanger technological development. Our interlocutor, however, notes that in this particular case, the public interest requires the protection of citizens` rights.
- The rules proposed by the EU are demanding and can create high costs for companies in the field of technology. According to some estimates, companies developing AI systems will have to budget 10-15% of their overall costs for legal and ethical compliance costs. These are large percentages when viewed from the perspective of companies, so they say that this will slow down innovation and prevent the competitiveness of the European technology sector compared to the Chinese and American sectors, where the development of AI will be regulated more liberally - says Krivokapic.
According to our interlocutor, these companies` arguments are valid, but we must be aware that we are currently facing AI as a phenomenon that we did not have at that level, so it is no longer a question of economic development, but of preserving society and its fundamental values.
- We will see who will win, whether more liberal approaches, where AI will be developed with much less attention to human rights, privacy, citizens` data and with a greater possibility for error and damage or a system that will be more restrictive, will require risk assessment and a lot of attention to how those risks to be reduced before the product is released on the market - the professor notes.
Decision-makers in states and companies have a key role in shaping the framework and regulations that will guide the development and application of artificial intelligence, Marko Pavlović believes. But the responsibility of engineers and experts involved in the development of artificial intelligence is also important in designing AI systems that are transparent, fair and safe for users, in order to minimize potential risks, he adds.
There are also educational institutions to educate future generations about ethical issues related to artificial intelligence, as well as lawyers and legal experts because understanding complex legal issues, such as responsibility for AI decisions or privacy protection, is crucial to ensure legal certainty in in the context of artificial intelligence.
Nevertheless, our interlocutor notes that each of us, regardless of role or profession, should be aware of the impact of AI technology on our lives and understand our rights and responsibilities in relation to it.
- It is also important to think carefully about what information we share online, because activating privacy settings on devices and applications helps to limit access to personal data. If we are talking about companies that develop and use AI, their task is to be transparent about how data is used and how decisions are made. Also, as many of the challenges of artificial intelligence are global, cooperation between countries and international organizations is essential. Harmonization of rules and standards contributes to consistent protection worldwide - says Marko Pavlovic.
Where is Serbia - From Ethical Guidelines to Law
And Serbia, if it wants to develop artificial intelligence, will have to harmonize legislation with EU regulations, so the adoption of the Law on Artificial Intelligence is also expected. There are already Ethical guidelines for the development, implementation and use of reliable and responsible artificial intelligence, which were adopted in March this year.
- The goal of the guidelines is to provide everyone who applies AI in their business with a framework for how they should do it. For example, transparency is one of the key ethical principles. Therefore, if someone uses AI in their work, they must inform the users that they are using artificial intelligence, as well as enable those who may feel the consequences of the application of AI to have human supervision, to have a person in the process who can verify the findings, decisions, recommendations that they arise from the work of AI - says Professor Krivokapic.
The law is the next step that Serbia, as a country in the process of European integration, will be obliged to adopt. But, as our interlocutor points out, it is not only the duty towards Europe, but towards our society and economy that is important.
- It seems to me that there is no obstacle that as soon as the law is passed in Europe, work on drafting a similar act will begin in Serbia, as we have already done for the Law on the Protection of Personal Data, which was passed in our country only a few months after its entry into force in the EU - reminds Krivokapic.
This is very important not only for the citizens but also for the economy, our interlocutor notes, because it is unlikely that anyone in Serbia will develop the AI industry for the local market.
- If domestic companies are interested in producing AI systems and exporting them, they will have to comply with the European framework, so it is good for our economy to have an identical legal framework in this area as Europe - says Krivokapic.
It develops too fast, but it is not (yet) all-powerful
Because of all the above, it often seems to us that we are unprepared for all the questions and dilemmas arising from the use of artificial intelligence and that its development is too fast. However, Marko Pavlović reminds that AI has the potential to improve many areas, including health, education, economy and research.
- Although the rapid development of artificial intelligence brings numerous doubts, with the active participation and engagement of all interested parties, we believe that we can create an environment in which AI technology serves the welfare of society and people - says Pavlovic.
He notes that artificial intelligence has indeed achieved remarkable successes in many fields, but there are still limits to its capabilities that we need to understand. As he says, AI is capable of efficiently analyzing huge amounts of data and drawing useful conclusions from it. This is essential in areas such as medical research, marketing and many others.
- It is also effective in automating routine tasks, as artificial intelligence can perform monotonous and repetitive tasks with precision and speed, such as word processing, warehouse management or performing regular machine maintenance. It is an exceptional tool for recognizing patterns in data and using that information to make decisions, such as facial recognition, fraud detection or voice recognition. Additionally, in many areas, AI systems provide useful information and suggestions that help people make better decisions, such as medical diagnosis or financial advice - Pavlovic explains.
However, there are limits when it comes to artificial intelligence, our interlocutor notes. Although it can generate songs, AI still lacks real creativity and emotional intelligence. Human artistry and the ability to empathize are areas in which artificial intelligence lags behind, Pavlovic adds.
-Also, AI has no ethics and moral understanding of its own. Decisions made by AI are based on programming and data without moral awareness or conscience. In addition, artificial intelligence can effectively solve certain types of problems, but it does not have the ability to think creatively and find completely new solutions. Although artificial intelligence is a powerful tool, it is not a substitute for human creativity, emotional intelligence and moral reasoning. Further scientific research and development in the field of artificial intelligence will help us to better understand its real potential and limitations, and will also allow us to use it responsibly in society - concludes our interlocutor.
Marija Dedić
Click here to see the entire Special Edition Newsletter
"DIGITAL TRANSFORMATION - Smart, game-changing solutions"
