What Is AI Ethics?

By Dr. Steven Wyre  |  08/31/2026

Because artificial intelligence (AI) technology is designed to augment or replace human intelligence, the field of AI ethics is crucial. The same biases that affect human judgment can also influence AI systems, potentially leading to harmful consequences for marginalized groups.

When considering the ethical principles involved in using artificial intelligence, government agencies, private-sector businesses, and education can create unique ethical situations. It is necessary to develop a framework about how much AI is too much and how its use should be documented.

AI ethics involve making positive, ethical use of AI tools while also minimizing any risks or negative outcomes. But the guidelines for the ethical use of AI are not always clear or even visible. Consequently, it can be challenging to know when someone has crossed that line into unethical use.

Questions about user safety and psychological dependence have become increasingly important in discussions about AI ethics. If Grok®, for instance, induces AI psychosis in a person who later commits suicide or causes harm to other humans, can we say it acted with malice or was immoral?

Part of this situation involves the notion of agency and subjective experiences. Another part involves what recourse humans have against AI systems or the people who build them.

AI Ethics and Machine Sentience

According to The Nobel Prize website, the godfather of AI, Geoffrey Hinton, won the 2024 Nobel Prize in physics for work “for foundational discoveries and inventions that enable machine learning with artificial neural networks.” It was Hinton’s and others’ efforts that opened the door to the creation of neural networks.

Neural networks are designed to enable machines to perform the same kinds of processing that a human brain uses to solve problems and communicate. Neural networks are involved in backpropagation and deep learning, which aids natural language processing.

But can current AI systems like Grok, Gemini®, ChatGPT®, and Perplexity® truly think, have human consciousness, and have subjective experiences? Hinton claims multimodal chatbots do.

Understanding AI Ethics and Moral Theories

When talking about ethics, there are three theories that each have their core principles. These theories are:

• Consequentialist – the consequences (usually desired consequences) matter most
• Deontological – a set of guidelines is more important than the consequences of the action
• Virtue – the kind of person one is rather than a set of rules or desired/actual consequences

All moral theories fit under at least one of these three types.

There are other theories, such as intuitionism, that could be promising in ethical AI development. When considering AI ethics, some moral positions would seem naturally intuitive to any rational person.

Unfortunately, many philosophers have argued against the possibility of objective moral facts, making it nearly impossible to identify a “self-evident” claim. It might be possible for someone to construct a more coherent argument for using intuitionism in evaluating AI ethical practices and ethical challenges.

Realistically, when considering AI bias, AI risks, AI regulations, and AI tools, knowing the right thing to do is not that difficult. There is, however, a sizable gap between knowing and following reasonable ethical standards.

In business, for example, business leaders must follow ethical guidelines that allow businesses to serve the needs of customers, employees, and other stakeholders while also making a profit. However, what falls under business ethics will depend on the desired business outcomes.

AI ethics principles are no different. No matter what values or guidelines are proffered, they fall under consequences, rules, or virtues.

Many countries and states are working to develop and implement AI regulations locally. Some AI regulations are already in place and many more are coming.

When discussing artificial intelligence, there are not only different applications but also many issues within each application. Take data centers for AI, for instance. What moral conundrum could data centers produce?

Maine has implemented legislation banning data centers. It is the first state to completely outlaw data storage facilities.

While a data center may create some jobs, the incredible costs of running it could raise power costs for other residents and increase pollution. The need to crank out more energy will increase the use of fossil fuels and raise carbon dioxide levels in the air. When considering responsible AI development, all of these considerations can be considered ethical matters.

AI Bias and Discrimination

In the race to be first, how many companies and researchers are stopping to ask important questions about AI ethics? When collecting massive amounts of personal data, what are companies doing to ensure data security and data privacy?

When Large Language Models (LLMs) incorporate data that reflects bias and racial perspectives, is it any wonder we see bias and racial perspectives in the answers it provides, the images it creates, and the job candidates it selects? When scouring the internet, are any AI companies seeking permission for the personal, copyrighted material they are hoovering up?

Discrimination is not new in evaluating resumes. With a majority of large companies using generative AI to recruit and screen applicants, it is no wonder human resources systems are exhibiting racial and gender bias.

In 2018, Amazon® faced backlash over its AI recruiting tool, which discriminated against female candidates because of biased training data. Amazon later abandoned the tool.

This case was an example of how AI systems can perpetuate historical bias and discrimination if the AI tool’s training data does not accurately represent a population. In turn, this use of AI can lead to unfair outcomes in hiring and facial recognition.

Many experts point to the environmental impact of AI applications. The high energy and water consumption required to train large AI models is an emerging ethical concern. While AI advocates proclaim the many benefits of artificial intelligence to humanity and promote human values, a cost-benefit analysis should be run.

As AI systems become more integrated into daily life, ethical questions arrive about accountability when users rely heavily on AI-generated advice or companionship. For instance, I have encountered rational humans who use AI chatbots for friendship or therapy, even though doing so could harm their mental health.

What do we do when the chatbot stops following orders? What are the ethical implications of generating deepfakes or producing sexualized images of real people without their consent?

It can be challenging to say that some AI chatbots are not as bad as others, but that raises other philosophical questions:

• Should developers place limits on AI chatbots that may encourage unhealthy emotional dependences or misleading interactions?
• What are business leaders doing to ensure transparency when AI tools are used in hiring and decision making?
• Where does human oversight end and machine learning take over?

These questions just scratch the surface of AI ethics. Any effort to examine the ethics of what AI creates would need to consider factors such as:

• Transparency and explainability
• Accountability
• Human rights abuse
• Increased security
• Increased hacking
• Corporate spying
• International terrorism
• Enhanced facial recognition
• Deception via deep fakes
• Job elimination

AI companies need to do more to promote fairness and display transparency. They should clearly explain what their AI software does and why.

AI Ethics, Governance, and Regulations

In terms of finding moral principles to guide actions and justify taking certain positions, some regulatory bodies have been at work:

• The Organisation for Economic Co-operation and Development (OECD) developed a set of value-based AI principles.
• UNESCO created recommendations on the ethics of artificial intelligence. It focuses on human rights, environmental sustainability, and inclusive, peaceful societies.
• The European Union (EU) passed the EU AI Act in 2024 as a form of AI governance.  

AI Ethics in Education

There is no shortage of organizations promoting ways for education administrators to harness the power of AI, whether for online learning or in-person learning. It might be a tremendous boon to have some AI algorithm tailor lessons to a person’s preferred learning style, suggest electives that are more suited to that person’s background, or assess a person’s ability to repay student loans.

However, are these uses of artificial intelligence really in the best interests of the student? If all a student sees are lessons tailored to an assumed “learning style,” doesn’t that approach prevent the student from learning in other ways?

I learned long ago that the smartest people cannot predict whether a student will succeed or fail. Do we honestly want to leave that type of decision up to AI-powered tools? Given AI’s propensity to hallucinate and engage in unethical practices, what guarantee do we have that AI software won’t set up the student to fail?

Embracing AI tools will lead to a growing gap between the haves and have-nots for students, faculty, and institutions. Schools are already touting just how far they will go in promoting AI, perhaps to their detriment.

While it seems likely that normal, rational adults should see these ethics issues and intuit the right thing to do, we may be dealing with less than rational adults. Any way you look at it, the greater good, the obligation of educators to teach, and the virtues of teaching point to the need to reevaluate the promotion of AI in education.

The Future of AI Ethics

Clearly, the onus is on AI developers to take responsibility for their products. Designing ethical principles for responsible AI use and development requires collaboration between industry actors, business leaders, and government representatives.

Unleashing AI irresponsibly and using it to harm others should be penalized. We are all culpable if we fail to ensure ethical AI governance, especially since we live in a world with rapid technological change.

We need rational minds to engage in open conversation about our approach to AI use and governance and apply various moral theories to justify a reasonable solution and a comprehensive framework for AI governance. We must consider the societal implications of our interactions with AI and develop workable guiding principles.

The Bachelor of Arts in Philosophy at APU

For students interested in studying philosophy, American Public University (APU) offers an online Bachelor of Arts in Philosophy. For this degree program, students can enroll in courses that cover an introduction to ethics, the philosophy of science, and logic. Other courses include critical thinking and research, analysis, and writing.

This degree program offers two concentrations to enable students to tailor their education to meet their professional aspirations:

• History of Philosophy
• Society and Justice

For more information about this degree, visit APU’s arts and humanities degree program page.

Grok® is a registered trademark of X.AI, LLC.
Gemini® is a registered trademark of Google, LLC.
ChatGPT® is a registered trademark of OpenAI OpCo, LLC.
Perplexity® is a registered trademark of Perplexity AI, Inc.
Amazon® is a registered trademark of Amazon Technologies, Inc.