Regulation for Artificial Intelligence: First Ever Agreement on the Ethics of AI
May. 12, 2022 • Nikita Saha
AUTHOR'S PROFILE: Nishka Kapoor, NALSAR University of Law, Hyderabad.
Introduction
Recently, United Nations Educational, Scientific and Cultural Organization (UNESCO) member states, i.e., 193 countries signed[i] the first-ever global agreement on the Ethics of Artificial Intelligence (AI). This agreement deals with the values and principles that need to be ensured for the development of AI.
Artificial Intelligence is a computer program that performs work that usually requires human intelligence. Artificial Intelligence is a major part of everyone’s daily life, it is useful in multiple ways and less time-consuming. However, at the same time, it also poses greater threats to the basic rights of an individual such as the right to freedom, right to privacy, and equality and so there is a need to make proper rules to eliminate the negative aspects of Artificial Intelligence for the growth and development of society.
Artificial Intelligence: Boon or Bane for the Humanity
Artificial Intelligence has benefitted society in innumerable ways, such as education, healthcare, law and justice, and social work. AI in many ways has played an integral role to uphold human rights in many areas of life, and during the time of pandemic[ii] that we are in now, we rely on technology more than ever. With the help of AI, even during the lockdown, we were able to carry on with our daily work smoothly. AI-enabled working of schools to jobs in an online setting as they would physically, and not only during the pandemic but also in normal conditions, AI has become a major of our life, the society, it is useful for doing small tasks such as buying clothes online to solve greater issues such as solving climate change effect.
However, there are many challenges that Artificial Intelligence poses in terms of the technicality and legality of developing the AI and the ethical challenges[iii] attached to it. One of the main concerns associated with Artificial Intelligence is Data Collection and Privacy. The legal issues mostly arise due to the technical errors and unethical usages of a person’s data, such as data repurposing and spillovers, unlawful surveillance, gender bias, and other issues related to cyber security, which pose a major threat to the rights of the citizens as well as the security of nations.
There are many instances where many big tech companies or organizations collect the data of users without their consent, and this also has a major societal impact, most time data is shared without the knowledge of users when they are using applications that are developed to collect and data of the user without asking them for the permission. And there are many scandals when AI systems collect the data of users in an unethical manner, one such major scandal was The Cambridge Analytica-Facebook Scandal[iv], where the firm Cambridge Analytica used Facebook for Ad campaigns to influence the psychological and social behaviours of their users before the 2016 US Presidential elections and was also accused of influencing the outcomes of the 2016 US Presidential Elections.
Another way through which AI collects data and invade the privacy of individuals is facial recognition, which is recognizing the human face using biometrics to map facial features from photograph or video. Facial recognition technology is used in identifying a person’s identity; however, this technology violates the privacy rights of the people and sometimes leads to wrong identification. Some of the recent cases where the organizations were banned from using facial recognition were Clearview AI[v], which was ordered to destroy all images and facial templates on the ground of privacy breaches, similarly, Facebook also shut down its feature of facial recognition[vi] and facial templates created in its system, because of the negativities associated with this feature. Facial Technology is not very safe as it might also become a threat to the security of a person.
Among other AI technologies that pose a threat to the security and privacy of the person is unlawful surveillance. Surveillance is supervising a person, carefully noticing, and keeping a track of their activities, this technology is often used to catch the wrongdoer or lawbreaker who has committed the crime, it is majorly used by police, the army, and anyone who is authorized by the law. One of the biggest surveillances carried out in the world is the one in China, ‘Mass Surveillance of China'[vii], where the government used facial recognition cameras and other surveillance systems to monitor tourist sites, airports, local residential areas, and so on, for region’s minorities Uighur population they deployed advanced surveillance systems to collect their data and detain them. The Chinese government by doing this is violating many major rights of its citizens and the minority community.
Creating artificial intelligence without a uniform system of development deepens the gender and ethnic inequality in society [viii]. This inequality is often based on the belief of the developer of the AI system and the stereotypes formed about gender and ethnicity. In the United States of America, three black people have been wrongfully arrested [ix]because of a false facial recognition match. In all three cases, the technology identified all three black people to be the wrongdoer, when they were not even aware of the incident that had happened, this incident shows the bias that is being shown by the AI towards black people. AI also poses significant threats to gender equality, the incident of this type of bias was with the AI tool developed by Amazon[x], which had to be taken down as it taught itself male candidates were preferable to women. The other reason why AI is considered discriminatory toward women, is that they are at a higher risk of losing their jobs because of job automation[xi], and most the women are not familiar with the technology, the primary factor being they are not given the right to educate themselves as men and so during such major changes in the working, that is shifting to digital mode of working they are often left behind.
Despite the many challenges that artificial intelligence poses, if regulated and operated properly it would be one of the greatest developments for humanity. Artificial Intelligence could play a vital role in protecting human rights by helping overcome major threats such as climate change, the rights of the person with disabilities, etc. that the world is facing.
It is often argued that AI has contributed to increased CO2 emissions[xii] because of the power-intensive machines that are used in its work. However, AI can play a vital role in reducing the effect of the climate crisis[xiii] and help in averting the adversities posed by it, as the system can learn and predict[xiv] the weather patterns, and through this information, it becomes easier to determine the magnitude of the climate change and in which areas of the world will it affect, and so accordingly it helps the experts to work on it and try to minimize the adversities that it might pose on the people residing in those areas. For example, meteorologists with the help of AI are very accurately able to a cyclone, its speed, and intensity and based on the predictions help people residing in the area to mitigate to safer places so that the lives of the people remain safe, and their loss is also minimized. The working of AI in the area of climate change has already begun, there are about 30 models related to climate change designed by AI[xv]and the Intergovernmental Panel on Climate Change[xvi] of the United Nations is utilizing those models.
Artificial Intelligence could also play a pivotal role[xvii] in upholding human rights, the technology, if operated properly, could enhance development in various areas of life, such as education, health, social work, law, and justice. There are many arguments about AI violating rights and going against human rights, but those are because of the lack of proper developmental policies, which leads to a lack of uniformity and miscarriage of rights by invading privacy and security.
There are regulations at the global level to regulate the technicalities and challenges associated with AI, such as the International Panel on Climate Change model to study the effect of AI[xviii] on people and economies around the world, this model was renamed Global Partnership on Artificial Intelligence[xix]and was more holistic in its approach as it focuses on the development of AI by Human Rights, Democratic Values and it also outlines ways to mitigate its negative effect and develop the trust of the people in technology.
Conclusion
There are very few countries that have made special provisions to regulate the development and working of AI, namely the US, EU[xx], Council of Europe[xxi], Canada, and Russia. However, the laws made on a global level or even at regional levels focus on the working of AI concerning the specific rights of the people. But, the agreement by UNESCO, is a progressive step toward development as has incorporated various aspects of the development of AI[xxii] and if implemented properly it would help in eliminating most of the challenges associated with AI.
Disclaimer- “The author undertakes that the work submitted is an original creation of the author. The author has not previously submitted the article for the purpose of publication. Any similarity with previously published content is not intentional. The author shall be personally liable for any infringement of intellectual property of any person, organization, government or institution”
REFERENCES
[i]https://en.unesco.org/news/unesco-member-states-adopt-first-ever-global-agreement-ethics-artificial-intelligence
[ii]https://www.oecd.org/coronavirus/policy-responses/using-artificial-intelligence-to-help-combat-covid-19-ae4c5c21/#:~:text=AI%20can%20also%20be%20employed,are%20speeding%20up%20medical%20diagnosis.
[iii]https://www.cloudmoyo.com/blog/ai-ml-automation/top-10-potential-ai-artificial-intelligence-problems/
[iv]https://www.cloudmoyo.com/blog/ai-ml-automation/top-10-potential-ai-artificial-intelligence-problems/
[v]https://www.theverge.com/2021/11/3/22761001/clearview-ai-facial-recognition-australia-breach-data-delete
[vi]https://www.nytimes.com/2021/11/02/technology/facebook-facial-recognition.html
[vii]https://fortune.com/2020/11/03/china-surveillance-system-backlash-worlds-largest/
[viii]https://www.mckinsey.com/featured-insights/asia-pacific/a-conversation-on-artificial-intelligence-and-gender-bias
[ix]https://www.nytimes.com/2020/12/29/technology/facial-recognition-misidentify-jail.html
[x]https://www.businessinsider.in/tech/why-its-totally-unsurprising-that-amazons-recruitment-ai-was-biased-against-women/articleshow/66192889.cms
[xi]https://en.unesco.org/system/files/artificial_intelligence_and_gender_equality.pdf
[xii]https://towardsdatascience.com/artificial-intelligence-has-an-enormous-carbon-footprint-239290ebffe
[xiii]https://www.techtarget.com/searchenterpriseai/feature/AI-and-climate-change-The-mixed-impact-of-machine-learning#:~:text=As%20explained%20by%20National%20Geographic,climate%20changes%20across%20the%20planet.
[xiv]https://www.techtarget.com/searchenterpriseai/feature/AI-and-climate-change-The-mixed-impact-of-machine-learning#:~:text=As%20explained%20by%20National%20Geographic,climate%20changes%20across%20the%20planet.
[xv]https://www.techtarget.com/searchenterpriseai/feature/AI-and-climate-change-The-mixed-impact-of-machine-learning
[xvi]https://www.ipcc.ch/
[xvii]https://datasociety.net/wp-content/uploads/2018/10/DataSociety_Governing_Artificial_Intelligence_Upholding_Human_Rights.pdf
[xviii]https://www.nationalgeographic.com/environment/article/artificial-intelligence-climate-change
[xix]https://gpai.ai/
[xx]https://law.stanford.edu/publications/eu-artificial-intelligence-act-the-european-approach-to-ai/#:~:text=On%2021%20April%202021%2C%20the,presented%20the%20Artificial%20Intelligence%20Act.&text=The%20EU%20AI%20Act%20sets,the%20territory%20of%20the%20EU.
[xxi]https://www.coe.int/en/web/artificial-intelligence
[xxii]https://news.un.org/en/story/2021/11/1106612