Third Way: The Dangers And Values Of AI In Democracy
Artificial Intelligence (AI) has been the subject of many science-fiction movies and television shows, usually in the form of an extremely realistic, humanoid, artificial being. This being inevitably becomes more powerful than its creators had expected and usually ends up turning against them, themes warning of the dangers of too much reliance on artificial intelligence and technology. However, in real life, AI hardly ever appears the way that it does in the movies. Instead, AI simply refers to an algorithm that analyzes data sets to perform tasks, which can take the form of things such as facial recognition software, self-driving cars, and Siri. Various forms of AI are currently being used by governments around the world and, while they have proven to be extremely beneficial, they have the power to be equally as harmful.
One example of Artificial Intelligence helping a government and a country is seen currently in Singapore. In the aftermath of COVID, Singapore’s tourism industry has drastically dropped, which was responsible for 4% of its economy. Hotels and companies within Singapore are now using AI to act as new security measures in order to help encourage people that traveling to Singapore is safe, even though there are still some COVID cases. The AI in use here is in the form of apps that act as crowd-monitoring to tell tourists which areas are more crowded, to help them safely socially-distance, as well as virtual check-ins and check-outs of hotels to reduce face-to-face contact. This will hopefully bring some tourists back to Singapore, which will help to jumpstart its economy again after all the losses that COVID caused.
However, not all countries are using AI in such a beneficial manner as Singapore. Surveillance systems and facial recognition technology have grown more advanced in the last couple of years and are now at play in dangerous ways. This is seen in China’s treatment of the Uighur Muslims in the Xinjiang province. Extensive surveillance systems and networks are in place throughout the province, both in the form of cameras that act as checkpoints to photograph the Uighurs, as well as apps on their phones that were forcibly installed to check for dissenting messages or any evidence that they were continuing to practice the Islamic religion. In this case, artificial intelligence is being used by the government to discriminate against and harm an entire group of people within their nation.
While this is not happening nearly to the same extent in the United States, facial recognition software in use in legal and justice systems has also proven to be discriminatory. There have been reported cases of this software wrongfully identifying and matching faces in crimes, particularly faces of black men, causing them to be wrongfully imprisoned for a crime that they did not commit. For one of these cases, the wrongful arrest of Robert Williams in Detroit based on facial recognition software, the ACLU has stated that this software violated his Fourth Amendment rights as well as his right to be treated equally regardless of his race. When the government allows this kind of software to be in use, with full knowledge of its discriminatory properties, the government is violating people of color’s right to equal treatment under the law. Until this software can be proven to be completely free of inaccurate matches, especially when it comes to people of color, it should not be used in legal and justice systems.
Yet, as seen in the case of Singapore, AI can be used to help governments and countries. One of the most important instances of AI use could be in medicine and healthcare. Artificial intelligence has already made rapid advances in the medicinal field, and will only make more in the upcoming years. AI can be used to predict the outcome of different medicines for cancer patients, as well as to predict people who are more at risk for certain diseases, such as type 2 diabetes. By making deep learning models, scientists can teach AI about the different diseases, which could potentially help to identify cancer cases in the first stage as they are just beginning, as opposed to in later stages when starting treatment might be too late.
In order to use AI in nations and governments in a beneficial way, there must be ethical regulations set in place. This is happening to a certain extent now, as the use of AI has become more prevalent, but it must continue to be greater regulated if it is not to infringe upon democratic principles. The House Ethics Committee has already imposed some regulations regarding certain aspects of AI among House members, especially as it pertains to deepfake videos. These are videos, often of either celebrities or political representatives, that have been altered by AI to intentionally mislead or spread misinformation, such as a video of Nancy Pelosi in 2019 that was edited to make her look drunk. By setting increased standards for house representatives as to what videos and photos they post, the House Ethics Committee is contributing to better uses of AI in the government.
The Organization for Economic Cooperation and Development (OECD), which the United States is a part of, has also set guidelines and principles regarding the use of AI In governments. These principles are designed in order to allow AI to be used to benefit people, while still making sure that human rights are not being violated during their usage. The second principle even includes a mention of democratic principles, stating that AI systems should be designed to respect democratic values and diversity. The guidelines and recommendations that accompany these principles focus on accessible and shareable AI, as well as newly developed policies, in order to ensure transparency and try to prevent AI from being used in less-democratic ways by infringing on people’s rights.
It is clear that, while AI has the potential to provide great benefits to nations and governments, it also comes with high costs of human rights and safety. It is currently being used in many ways that are counter to democratic principles, especially the democratic values instilled in the United States in the form of equal protection as well as free speech. However, more guidelines and regulations are being developed almost at the same rate as the development of AI. Hopefully, with these guidelines instilled, AI will be able to be used in ways that are more beneficial and further democratic principles than those who harm it.