AI Ethics

 

 

As human beings, we adhere to moral principles to guide our actions, ensuring our lives are comfortable and characterised by good manners and behaviour. These principles, which help us distinguish right from wrong, are known as ethics. Similarly, AI systems and tools are governed by their own set of ethical principles, referred to as AI ethics.

 

​AI ethics refers to the basic principles of AI system design that use the good code of conduct and produces the results.

 

In other words ethics means what is right and unethical means what is wrong.

 

Let’s take a look at the given ethical scenarios.

Ethical Scenarios - I

 

Imagine a situation where you are a doctor at a busy hospital. You have to analyze numerous medical records and patient histories every day, which can be overwhelming. You find an AI tool that can help you review these records quickly and provide recommendations for treatment.

​

AI ethics refers to the basic principles of AI system design that use the good code of conduct and produce the results. 

​

In other words ethics means what is right and unethical means what is wrong. 

 

Let’s take a look at the given ethical scenarios.

​

Ethical Scenario – I

 

Imagine a situation where you are a doctor at a busy hospital. You have to analyze numerous medical records and patient histories every day, which can be overwhelming. You find an AI tool that can help you review these records quickly and provide recommendations for treatment.

 

Think: "Learners, imagine yourselves in this scenario before moving on to the discussion questions."

 

Let’s Discuss:

1. Would you use the AI tool to analyse the medical records and provide recommendations?

2. Why would you do that?

3. What will be the advantages and disadvantages of using the AI tool in this context?

4. Can you think of any challenges which the AI tool might face?

 

 

Ethical Scenario – II

​

Imagine a situation where you are a journalist for a major news outlet. You need to write multiple articles each week and sometimes struggle to keep up with the workload. You find an AI tool that can help you draft

articles based on brief inputs and data you provide.

 

Think: "Learners, imagine yourselves in this scenario before moving on to the discussion questions."

 

Let’s Discuss:

1. Would you use the AI tool to draft your articles?

2. Why would you do that?

3. What will be the advantages and disadvantages of using the AI tool in this context?

4. Can you think of any challenges which the AI tool might face?

 

Understanding AI Ethics

​

AI ethics refers to the fundamental principles and guidelines that govern the development and deployment of AI systems to ensure they operate in a manner that is ethical and aligned with human values. These principles cover various aspects, including fairness, accountability, transparency, and privacy.

​

• Fairness: Making sure AI systems don't keep or increase unfair biases and that their decisions treat everyone equally.

• Accountability: Making it clear who is responsible for what AI systems do, so developers and users can be held responsible.

• Transparency: Making AI operations understandable and explainable to users, so they know how decisions are made.

• Privacy: Protecting individuals' data and ensuring that AI systems handle personal information responsibly.

 

Morals vs Ethics

​

​

 

 

 

 

 

 

Fun activity:

Purpose: Use Moral Machine Platform to exercise the morality of persons.

 

Moral Machine is a platform for gathering a human perspective on moral decisions made by artificial intelligence, such as self-driving cars.

At the end, you will be able to see how their responses compare with other people.

 

Activity Guidelines:

To perform the activity:

Go to this https://www.moralmachine.net

​

Let’s summarise: 

• The results will tell you which characters you preferred over the others. 

• Saving more lives matters to you. When given a choice, you would prefer to save as many people as you can. 

• It does not matter to you much if a person obeys the law or not when it comes to saving people. 

• You will also get to know what beliefs you value with the choices you make in the game. 

• You prefer protecting passengers, instead of pedestrians more. 

• When an equal number of people are getting hurt, you prefer to not be a part of the consequences, and you do not intervene.

Data management AI Ethics

 

These all data Google is using for the exchange of the details whenever such apps are used in your mobile. Sometimes it improves user experience as well. You will get the information from your app as per your previous transactions and actions on the app.

There are two main categories of AI ethics:

​

1.  AI Bias

​

AI bias occurs when an AI system makes unfair decisions because the data it was trained on is biased or unbalanced. For example, if a hiring AI is trained mostly on resumes from one gender or ethnic group, it might favor candidates from that group over others, even if they are equally qualified. It’s crucial to ensure that AI systems are designed to treat everyone fairly and equally.

​​

2.  AI Access

​

AI access refers to who has the ability to use AI technology and who might be left out. For instance, advanced AI healthcare tools might be available in wealthy countries but not in poorer regions, which means people in those areas might not get the same benefits from AI. This can lead to inequalities, so it’s important to make sure that everyone, no matter where they are, can benefit from AI equally.

​

​

Components of a good AI Policy

​

Good policy refers to the concerns, which should be in consumers' favour. 

The following topics can be considered as good AI policy: 

​​

Transparent System

 

A transparent system refers to the guideline and system purposes should be very clear to its users. While collecting data the purpose and the detailed guide about what to be done with the data should be known to the users.

​

Right of data collection

 

When the data is collected by the AI system it must be right to the data which the system is collecting. Without the collection of data, it cannot take the right decision for the user.

​

Freedom of leaving system

​

The user must have the freedom to leave the system. After using such system if user want to leave the system, the freedom should be given to the users.

​

Design

​

The system should be designed in such a manner that the data collection and purpose should be limited. This helps the users to stay and use the system in a good manner. The interface itself provides such controls to users.

​

Data Deletion

​

When the user leave the system, his data should be deleted. Or sometimes user requests to delete their data it should be provided in the system itself. In next section of AI ethics class 9 we will discuss another category of AI ethics.

​

Scenario III

Let’s learn some more about Jack:

​

He is an average middle school student. His school recently started using an AI-based essay grading system. The system takes in an essay and assigns grades after evaluation. Jack is worried that he scored a bad grade, even though he wrote a really good essay.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Let’s discuss 

• What do you think happened here? 

​

• Why did the AI evaluate Jack’s essay incorrectly and he received a lesser grade than expected? 

​​​

​

Real-World Examples of AI Ethics

​

To better understand AI ethics, let's explore some real-world examples where ethical considerations in AI development and deployment are evident.

​

Example 1: AI in Healthcare

​

AI tools are increasingly used in healthcare for diagnostics, treatment recommendations, and patient monitoring. While these tools have the potential to improve patient outcomes, they also raise ethical concerns:

• Bias in Medical Diagnosis: If an AI system is trained on data that lacks diversity, it may produce biased results that do not accurately represent all patient populations. For example, an AI tool trained primarily on data from one demographic may not perform well for patients from different backgrounds, leading to misdiagnosis or ineffective treatments.

​

• Privacy and Data Security: Healthcare data is highly sensitive. Ensuring that AI systems comply with strict privacy regulations (such as HIPAA in the United States) is crucial to protect patient information from misuse or breaches.

 

 

Case Study: IBM Watson for Oncology

​

IBM Watson for Oncology was developed to assist oncologists in diagnosing and treating cancer. While the AI system showed promise, it also faced criticism for producing recommendations that were sometimes inconsistent with expert opinions. This highlighted the importance of ensuring AI tools are rigorously validated and used as support systems rather than replacements for human expertise.

 

Example 2: AI in Autonomous Vehicles

​

Autonomous vehicles (AVs) promise to revolutionize transportation, but they also present ethical challenges:

​

• Safety and Accountability: Ensuring the safety of AVs is paramount. Developers must rigorously test these systems and establish clear protocols for accountability in case of accidents.

 

• Moral Decision-Making: AVs may face situations requiring moral decisions, such as choosing between two harmful outcomes in an unavoidable accident. Designing algorithms that make ethical choices in these scenarios is a complex challenge.

 

Case Study: The Trolley Problem and Autonomous Vehicles

​

The Trolley Problem is a famous ethical dilemma that questions whether to divert a runaway trolley to a track where it will kill one person instead of five. Autonomous vehicles may face similar dilemmas, forcing developers to encode ethical decision-making into their algorithms. This raises questions about who decides the ethical priorities and how these decisions are implemented in practice.

 

What are the principles of AI Ethics? 

​

Identifying the principles 

​

To make AI better, we need to identify the factors responsible for it. The following principles in AI Ethics affect the quality of AI solutions:

​

​

​

​

​

​

​

​

​

​

​

​

​

Let’s look at the AI Ethics principles in detail: 

​

Human Rights 

​

1. When building AI solutions, we need to ensure that they follow human rights:

• Does your AI take away Freedom? 

• Does your AI discriminate against People? 

• Does your AI deprive people of jobs? 

• What are some other human rights which need to be protected when it comes to AI? 

​​

Bias 

​

• Bias (partiality or preference for one over the other) often comes from the collected data. The bias in training data also appears in the results:

• Does your data equally represent all the sections of the included populations? 

• Will your AI learn to discriminate against certain groups of people? 

• Does your AI exclude some people? 

• What are some other biases that might appear in an AI? 

​​

Privacy 

​

We need to have rules which keep our individual and private data safe. Here are a few things that you should take care of: 

• Does your AI collect personal data from people? 

• What does it do with the data? 

• Does your AI let people know about the data that it is collecting for its use? 

• Will your AI ensure a person’s safety? Or will it compromise it?

• What are some other ways in which AI can breach someone’s privacy?  

 

Inclusion 

​

1. AI MUST NOT discriminate against a particular group of population, causing them any kind of disadvantage:

• Does your AI leave out any person or a group? 

• Is a rich person and a poor person benefitted equally from your AI? 

• How easy is it to use your AI? 

• Who does your AI help? 

• How can we make AI more inclusive?

​​

​​​​​​​​​​​​​​​​​​​​​​

Anti-Bullying with AI

 

Before we were so connected by technology, bullying was most frequently identified in school settings. But Bullying is not just for kids. Harassment and hate speech exists in the workplace and socially among adults. Online activity, such as social media, texting, and gaming, provides constant contact and further reach with our peers, colleagues, friends, and even strangers. Masked behind a keypad or keyboard, some people find the courage to publish disparaging and hateful posts that they might not have carried out in a face-to-face situation. And the problem for many young victims is that they do not report such instances.

​

Can AI help?

Yes, absolutely.

 

You can train an AI application to recognize texts/posts and match them to a particular output. No coding as such is required. You can enter multiple expression, including single words and sentences.

Referring to the image given below, The instructions are on the right side of the menu and help you to:

​

• Create the training data

• Train the Artificial Neural Network (ANN) through the Start Learning button

• Test the ANN by entering short sentences and phrases

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

On entering the inputs, Artificial Neural Network (ANN) will be constructed in real time so you can analysis how AI is giving the output in real life.

​

Questions/Answers

​​

What do you understand by Ethics?

• The concerns related to good and ideal behaviour or manners and moral principles are known as ethics.

• In other words, ethics refers to what is right and what is wrong then teaches moral conduct.

​

What are the moral issues related to self-driving cars?

• Self-driving cars are totally under the control of AI.

• For example, an animal is crossing your way suddenly hit by a self-driving car. The animal got hurt.

• Who is responsible for this accident?

• The following are some ethical concerns:

  • The developer of the car goes through all dilemmas while developing the car’s algorithm.

  • The morality of the developer gets transferred into the machine as what according to him/her is the right priority and hence would be the selection made by the machine.

  • Every person has their own perspective and he/she takes the decision according to their moralities.

​

Why ethical practices should be followed while developing solutions using AI?

• Ethical practices are very important while developing solutions using AI because the solution is made to solve the problem or reduce the complexity of the problem.

• If ethical practices are not followed properly then the solution might create more problems and that leads to more problems.

• Even unethical practices increase risk and damage to the surroundings. Which is not a good cause for the problem solution.

​

​​