Ethics and the Use of Artificial Intelligence
*Generative AI is used in the writing of this post
What is Ethics?
Ethics, in its broadest sense, refers to the set of principles and values that guide human actions, determining what is considered right or wrong, good or bad. In general, ethics helps individuals make responsible decisions based on what is considered just or morally acceptable.
When ethics is applied in the digital realm, and specifically in the field of Artificial Intelligence (AI), crucial questions arise: How should we use technologies to avoid harming society? What norms should govern the behavior of intelligent systems that interact with us more and more deeply? Ethics in AI is not just about creating innovative technologies but also ensuring that these technologies operate in a way that is equitable and respectful of human rights.
What are Ethical Dilemmas?
An ethical dilemma arises when a person or group must make a decision, but there is no clear option that is entirely right or wrong. Ethical dilemmas generally involve a conflict between two or more moral principles, and the final decision may depend on the values that are prioritized in that situation.
In the context of AI, ethical dilemmas arise when decisions made by automated systems significantly affect human lives. For example, if an AI system must decide who should be granted a loan, but the system relies on historical data that perpetuates racial or gender biases, an ethical dilemma arises about whether the technology is acting justly.
Most Common Ethical Dilemmas in Digital Ethics
- Privacy: The collection and analysis of personal data by AI raises significant concerns about privacy. AI technologies, such as virtual assistants or social media platforms, collect data about our behaviors, preferences, and interactions. While this allows for personalized services, it can also lead to abuses. A common ethical dilemma is how to balance the utility of data collection with people’s right to keep their personal information private. Imagine a company uses a user’s browsing data to personalize ads, but the user is not fully aware of how their data is being used. Is it ethical for the company to collect and use that information without explicit consent?
- Transparency: Transparency in AI algorithms and decision-making is another ethical dilemma. Often, AI systems operate as “black boxes,” meaning it’s difficult to understand how decisions are made. In fields like criminal justice or healthcare, it is crucial to know how and why an automated decision is made. If an AI system decides not to grant a loan to someone based on certain data in their financial profile, how can people know if that decision is fair or if the system is biased in any way?
- Responsibility: Who is responsible if an AI makes an error that causes harm? This ethical dilemma arises when automated systems make decisions that directly impact people’s lives. When AI makes a mistake, like misclassifying medical images, it is difficult to determine whether the blame lies with the system’s creator, the user who implemented it, or the technology itself. If an autonomous vehicle’s AI system causes an accident, who is responsible: the vehicle manufacturer, the software developer, or the person driving the vehicle?
- Inequality: AI has the potential to exacerbate existing inequalities, whether by amplifying biases in the data or being inaccessible to certain groups. The ethical dilemma here is how to prevent automated systems from perpetuating social, economic, or racial biases that already exist in society. If an AI system used in hiring discriminates against candidates from certain ethnic backgrounds due to the historical data used to train the model, an ethical question arises about justice and equal opportunity.
Generative Artificial Intelligence
Generative Artificial Intelligence is a subcategory within AI that has the ability to create original content, such as text, images, music, and more. Well-known generative systems include models like GPT-3 or DALL·E, which can produce coherent texts or completely new images based on an input request.
These advancements, while exciting, also present significant ethical dilemmas. For example, generative systems can be used to create false or manipulative content that is difficult to distinguish from genuine content. AI could generate fake news, manipulated images, or even deepfake videos, which jeopardize public trust and the integrity of information.
A generative AI model could create a fake news article that looks real, deceiving people who read it. Is the AI creator responsible for the damage caused?
The Phenomenon of AI Hallucinations
“Hallucination” in AI refers to a phenomenon where generative models produce incorrect, invented, or factually untrue content. This can be particularly problematic in fields like medicine or law, where errors can have serious consequences.
An AI system helping doctors diagnose diseases might generate an incorrect treatment recommendation based on faulty data, which could harm a patient’s health.
Hallucinations in AI models are difficult to predict and can be harmful if not managed properly. Those using these systems must be aware of the limitations and ensure that any information generated is verified before making decisions based on it.
So, Should We Stop Using AI?
The question of whether we should stop using Artificial Intelligence, especially generative systems, is valid and should be asked seriously. However, the answer is not simple. AI has immense potential to improve our lives and practices in areas like education, healthcare, the arts, and more. The problem is not with the use of AI itself, but with how we use it and the ethical principles we follow.
For example, what would happen if this text you are reading now had been partially generated by AI? Would it be an error or an ethical problem? Let’s see…
If this text was generated by AI, but the information in it has been verified and checked, and it is clearly stated that the content was assisted by AI, would this be an error or an ethical problem? In this case, not necessarily. Using AI as a tool does not imply plagiarism or misinformation, as long as certain ethical standards are met. It is important that any use of AI is transparent, honest, and ensures the accuracy of the information.
If this text was partially generated by AI, and it was clearly stated that the technology was used transparently, then it would meet ethical standards. What is crucial is acknowledging the use of AI and maintaining the integrity of the information. The ethical problem arises when the use of AI is hidden, or incorrect information is presented without proper verification.
Steps for Using AI Ethically as an Assistant
- Transparency: Always clearly state that you are using AI to assist in content creation. Users should be aware of the AI’s participation in the process.
- Verification of Information: Ensure that all information generated by AI is checked and validated. AI systems can make mistakes and generate incorrect information, so it is essential to corroborate facts and avoid spreading false information.
- Responsibility: As an AI user, it is your responsibility to ensure that the generated content is ethical, accurate, and does no harm. Do not use AI to create content that could mislead, manipulate people, or violate copyrights.
- Credits and Citations: If AI has assisted in creating a text, the proper credits should be given. If the content was partially generated by AI, this should be clearly stated, just as any other source would be cited.
- Avoiding Plagiarism: Do not present AI-generated work as entirely original or as if it was created without assistance. This ensures no plagiarism is committed, and the content remains an honest piece of work.
- Critical Reflection: When using AI, critically reflect on how it is affecting the content creation process. Are you delegating too much to the technology? Are you making responsible decisions about how and when to use it? Always maintain control over the final product.
Below is a table with examples of situations where the use of AI in academic tasks could be ethical or unethical, to help teachers and students reflect on its use.
| Situation | When is it ethical? | When is it unethical? |
|---|---|---|
| Writing an essay for a university assignment | – Use AI as an assistant to generate ideas or summaries. – Verify and complement AI-generated information with reliable sources. – Declare the use of AI in the process. | – Use AI to write the entire essay and present it as your own without mentioning assistance. – Not verify or cite the information generated by AI. |
| Creating academic presentations | – Use AI to generate images or suggest slide designs. – Use AI to organize ideas and structure content. | – Copy and paste AI-generated content without modification or personal analysis. – Not give credit for the use of AI tools. |
| Developing a research plan | – Use AI to explore relevant literature or generate initial research questions. – Check and expand AI’s suggestions with valid academic sources. | – Allow AI to generate the entire research plan without the student making in-depth analysis or proposing critical approaches. |
| Solving mathematical or scientific problems | – Use AI to check solutions or receive suggestions on problem-solving approaches. – Reflect and understand the process behind the generated solutions. | – Use AI to solve all problems and submit the results without understanding or explaining the process. |
| Creating exercises or educational materials | – Use AI to generate examples, activities, or questions based on the curriculum. – Customize AI-generated educational materials to adapt them to the group’s needs. | – Copy exercises or materials directly without customizing or adapting them to the specific context of the classroom or task. |
How to Use AI Ethically in Education
- Use as an Assistant: AI should be used as a complementary tool to facilitate the learning and content creation process, but it should never replace the intellectual work of the student or teacher.
- Verification and Complementation: All information or content generated by AI should be verified and complemented with reliable sources. Students should apply critical thinking to what AI provides them.
- Transparent Declaration: Students and teachers should be transparent about the use of AI in their work. If AI has assisted in content creation, this should be clearly stated, just as any other source would be cited.
- Development of Critical Skills: While AI can assist with tasks like writing or generating ideas, students should use these tools in ways that foster their learning and skill development, not as a way to avoid intellectual work.
- Use for Translations: AI can be a very useful tool for translating texts from one language to another. However, it is important to remember that automatic translations may not be perfect. Students should check the translation and adjust the content to ensure it accurately reflects the original meaning and context. Instead of relying solely on AI, it is advisable to review the translation and, if possible, compare it with other sources or ask an expert for help.
- Verification of Grammatical Issues: Students can use AI to check and correct grammatical errors in their texts. AI can help identify errors in agreement, punctuation, and other grammatical aspects that may be overlooked. However, it is essential that the student also manually review the work to ensure that the context and tone are appropriate. AI does not always understand nuances or the author’s style.
- Adjusting Reading Level Based on the User: AI can also be useful for adjusting the difficulty level of texts according to the student’s needs. For example, it can generate simplified summaries of complex texts or adjust the complexity of content to suit different levels of language proficiency. This allows students to access materials that are easier to understand and facilitates gradual learning. However, it is important for teachers to review these materials to ensure that academic rigor is maintained and key concepts are not oversimplified.
The ethical use of Artificial Intelligence in education depends on how it is integrated into the teaching-learning process. Used responsibly, AI can be a powerful tool to help teachers and students maximize their productivity and creativity without compromising academic integrity. Through transparency, information verification, critical thinking, and the responsible application of tools such as translation, grammatical correction, and text leveling, we can ensure that AI becomes an ally in education.
