The European Commission's Communication on Artificial Intelligence defines AI as “Artificial intelligence refers to systems that display intelligent behavior by analyzing their environment and taking action with some degree of autonomy to achieve specific goals.''
An artificial intelligence system uses predefined algorithms to turn input data into outputs in order to make decisions by following a set of instructions.
The term 'ethics' means moral philosophy. It is the science of proper behavior. It helps us to discern between what is good and what is bad for individuals as well as societies. AI ethics is the branch of ethics that deals with the moral issues during the creation and application of artificial intelligence. Or the responsible use of technology. It contains a set of guiding rules and principles that stakeholders (from engineers to government officials) use to ensure AI technologies are developed and used ethically. In other words, using AI in a safe, ecological, and more humane way.
As we know, the influence of AI is increasing day by day in almost all aspects of life, which raises concerns in privacy, fairness, accountability, and safety. The past five years witnessed the shifting of AI ethics discussion from academic coverage into public as well as political arenas.
Every day, increase the scope of AI technologies. It gets more powerful day by day. Each task has its own AI system, like;
* Photo recognition AI
* Self-driving AI
*Generative AI
* Face recognition AI
* Natural language processing AI
Rather than human-based instructions, many AI algorithms train themselves by analyzing the data they receive. The question that arises from here is whether all the AI outputs are trustworthy. If anyone feeds biased or inexact data, there emerges a chance of harmful outcomes. As an example, Google facial recognition software identified African Americans as gorillas. It had a bias against African Americans based on the data it learned.
Ethical issues related to AI implementation
1) AI Bias
AI is not a rational being like humans. It doesn't think the way as we do. Because of the lack of ethics and values, a number of risks end up in AI-biased actions. Any bias can transfer from the developer to the machine either at the time of designing the algorithms or at the time of giving the data to the AI model. A report from Pro Publica says that in the US, the AI system was biased in that it provided longer sentences to black defenders as compared to white defenders for the same crime. AI algorithms can become racist, sexiest, and unfair. Such biased algorithms may provide unjust or unbalanced results, which may especially harm marginalized or underrepresented individuals and groups.
2) Privacy issues
The information that we put into different applications, websites, and systems is stored in the database. The data gathered in this way can sometimes be used for malfunctions, which enhances risk possibilities.AI systems used in marking attendance or entry of employees by collecting facial and demographic characteristics, collection of patient data in health care, etc. could sometimes affect privacy concerns. Data plays a crucial role in every aspect of artificial intelligence. In recent times, the amount of data collected via gadgets, applications, and social media platforms is large, which accelerates AI activities.
3) Transparency
Transparency is essential to the decision-making process. AI systems are often opaque, making it difficult to hold them accountable.
4) Unemployment
As AI is improving our standard of living and assisting us in all our daily tasks. It also led to the increase of the unemployment rate world-wide. In future,
Human drivers will be replaced by self-driving cars.
* AI robots can be used for treating patients in hospitals (now in China).
*Chatbots are conquering the roles of people in customer support services.
* There will be a lesser need for receptionists in hotels due to auto check-ins.
Some are of the opinion that AI will reduce jobs; others have the opinion that AI will increase job opportunities. Some claim that there will be no effect on the job after the arrival of AI.
5) Inequality
The influence of artificial intelligence somehow creates inequality in society. Economic inequality may worsen. It reshapes the labor market. Human labor loses its values. It may create discrimination between people who can use software and those who are unable to.
6) AI access
AI is used in more advanced machines and technologies.AI can access only a few people.The elite class of society has the potential to enjoy the new possibilities of AI. It will create a gap between people and get widen with the evolution of AI technology.
7) Problem of facts and their interpretation
AI systems are unable to take logical decisions.All the outcomes are based on the data we entered. It doesn't know the reason behind the actions they took.
8) Security
AI systems can be used for both good and bad purposes. One of the most prominent fields in which AI ethics will come into play and where the transparency and accountability of AI systems will be crucial is defense. Bringing AI into defense.
9) Law
Using historical crime data as input in various cases creates discrimination among people. In many cases,AI makes decisions based on race and ethnicity as per past data experiences.
10) AI for kids
In this fast-growing world, children have tremendous opportunities to deal with AI applications. They can easily access all types of information. It may sometimes end up in dealing with unwanted things. Proper awareness should be given to them for the ethical usage of AI.
How to implement AI ethics
In order to implement AI ethics, we should implement a set of guidelines or rules that we want to
follow whenever we are creating or dealing with AI systems.
Elements of a good AI policy
* Transparency in AI systems: making AI systems understandable and their decisions explainable
* Accountability
* Fairness: equitable treatment of individuals or group of individuals
* Data for AI must be limited by design.
* Consumers may leave the AI system when they want.
AI initiatives developed today have inspired guidelines and frameworks that are recommended but not compulsory. Developing ethical standards for data processing using AI can be an important step towards setting up a legal foundation for ethical AI. AI is giving us an endless number of new applications that benefit all aspects of our lives. As AI data-driven technologies continue to develop, the threats and opportunities are also increasing. There should be enough representation of human rights while developing AI software. AI harm in various fields should be examined with proper policies and strategies.
Make sure AI is fair and doesn't treat people unethically based on gender, class, ethnic background, race, or religion. We should ensure AI is helping us, not hurting. A strong AI code of ethics should also be created at the international level to avoid biasing and to ensure the privacy of users.
