Exam Preview for 2375: The Ethics of Artificial Intelligence in Soc

The Ethics of Artificial Intelligence in Social Work Practice

Please note: exam questions are subject to change.

1. Which of the following examples demonstrates an “ethically responsible” use of artificial intelligence in social work practice? A social worker feeds client data into an AI program and relies fully upon the AI program’s clinical assessment of the client. A social worker uploads private client information into an AI program that sounds responsible but does not have a Business Associate’s Agreement as per the Health Insurance Portability and Accountability Act. A social worker uses an AI tool that maintains data on the social worker’s secure computer, ensuring that the data is not shared with the AI provider. A social worker refers a client to use an AI problem-solving program without checking whether the program is safe or effective.
2. What is one of the primary advantages of using an artificial intelligence program to summarize a meeting? AI is always more accurate at summarizing meetings than humans. AI can improve efficiency by providing a draft that the social worker can then check for accuracy. AI is able to accurately identify the psychological disorders underpinning each person’s responses during the meeting. Social workers do not have to be concerned about privacy if an AI program is summarizing a meeting.
3. A social worker uses an AI program to help conduct the intake process for new clients. Ethically, the social workers should: Tell the client, in advance, that the social worker is conducting the intake without AI. Only tell the client that AI is being used do conduct the intake if the client specifically asks. Inform the client that the intake process will be conducted with AI, unless the client would prefer another process for intake. Advise the client that it is inappropriate to ask whether the social worker is using AI.
4. A social worker is helping two clients with communication in the aftermath of a messy divorce. The social worker would like them to use an AI program that can review their emails to one another to provide them with guidance on how to ensure their wording is constructive (not rude or inflammatory). Ethically, the social worker should inform them that: AI’s suggested wording is always better than their own wording. They must use AI’s wording, even if it seems problematic. They should ignore whatever AI suggests because it’s likely to be more problematic than their own wording. They should consider the wording offered by AI, but they should take responsibility for the words that they ultimately choose to use in their emails.
5. Which of the following is an “unethical” use of AI in social work practice? Offering clients access to an AI program that can help them identify options for problem-solving. Requiring clients to use an AI program to generate options even if they don’t want to and don’t understand the program. Helping clients use AI to generate options and then going through the options with them to see if any of them are helpful. Letting clients know the potential benefits and risks of using AI to generate options.
6. Which of the following is an “unethical” use of AI in social work practice? A social worker uses AI to grammar-check an article and acknowledges the use of AI when submitting the article for publication. A social worker uses AI to assess a client’s mental health, knowing that the AI program is inaccurate and has fundamental biases. A social worker uses an effective and accurate AI program to assist with analysis of data as part of a program evaluation. A social worker provides a client with a choice of AI programs to assist with stress management strategies, explaining the pros and cons of each program.
7. A social worker is planning to go on a long vacation. Which of the following is an “unethical” use of AI as the social worker’s plan to help clients while away? The social worker instructs clients to that they have the option of asking a social robot for basic information and assistance, in addition to contacting a professional colleague for more in-depth and personalized assistance. The social worker explains the pros and cons of using a social robot for assistance, as well as alternative sources of help. The social worker provides a client with three options for help while the social worker is on vacation, including the possibility of using a social robot. The social worker tells clients that they can trust the worker’s social robot to be more accurate, empathic, and helpful than a human social worker.
8. A social worker would like to use AI to gather information and provide screening assessments, but it concerned about how to protect client confidentiality. Which of the following options is best for protecting client confidentiality? Using an AI program that has a Business Associate Agreement and is compliant with the Health Insurance Portability and Accountability Act privacy provisions. Using an AI program that stores all the client data on the Internet and makes the data available to the developer for continuous improvement of the AI program. Using an AI program that is free and available to anyone who creates a user account with the AI developer. Using an AI program that informs clients that the program may be based on algorithms with unintentional biases.
9. A social worker wants to use an AI program to assist with diagnosing mental health conditions. The social worker is not sure about the accuracy of the AI program’s ability to diagnose. Which of the following options is most ethical? The social worker should fully trust the AI program’s diagnoses without question. The social worker should advise the client that the AI program can diagnose more accurately than a human mental health professional. The social worker may use the AI program as a tool for diagnosing, but the social worker should verify the program’s diagnosis using evidence-based approaches for diagnosing. The social worker should tell the client that the AI program is replacing human social workers and that there is nothing else that the social worker can do.
10. Which of the following methods is the most effective approach for monitoring and assessing the effectiveness of an automated social intervention that uses AI? Ask the AI program to assess its level of effectiveness. Monitor the use of the program, gathering information and feedback from the client, and assess the effectiveness and risks of the program on an ongoing basis. Try the AI program for a couple of years before conducting an evaluation of its effectiveness. Develop a study focusing on the positive outcomes of the use of AI and avoid monitoring for possible risks or harms.