Artificial Intelligence (AI) Resources

As AI becomes more prominent in higher education, a variety of concerns have arisen, specifically with Generative A.I.  Below are the most common concerns attached to AI in higher education:

• Lack of Transparency
  • Since generative AI is still so new, these models can be unpredictable, and companies do not always fully understand the inner workings of these programs.  More studies will need to be conducted on how these programs work and how institutions can better understand these programs.
• Accuracy
  • Studies have shown that some Generative AI programs, such as ChatGPT, can generate false and inaccurate information.  Specifically, students can be vulnerable to disinformation, believing that the AI produces factual information.  Institutions must be aware of this and determine its accuracy, appropriateness, and usefulness before relying on or distributing information made by AI.
• Bias
  • AI systems learn from the data that they receive and are trained on.  If the data contains any bias, the AI will produce material containing this bias and could spread misinformation.  This concern becomes more apparent with potential bias surrounding political and philosophical bias, and the AI may produce responses that do not align with the values of the institution.  Institutions need to develop policies to detect and regulate biased AI outcomes that align with the organization's policies and legal standards.
• Cybersecurity and Privacy
  • When interacting with Generative AI tools, input may contain private and sensitive information.  This information may be stored and analyzed, which puts the user at risk of having their confidential information distributed.  While using AI systems, users should refrain from inputting sensitive information, such as private conversations, personally identifiable information, health records, and financial records.  Institutions must also prepare for their malicious use of Generative AI systems and cyber and fraud attacks by putting controls in place.
• Plagiarism and Academic Integrity
  • Since generative AI is easily accessible, students may use these programs to complete homework or testing assignments.  AI can also generate essays based on the prompt provided by the user.  Although cheating has always been an ethical concern in higher education, discerning between original and AI-generated work poses another obstacle.  Although there are some programs to detect AI in writing, such as Turnitin, they are not very accurate and can falsely accuse students of plagiarism.  Another issue that arises is students using generative AI for their work may hinder their ability to develop critical skills and learn the content of the courses.  Both students and teaching professionals will need to learn more about and understand the inner workings of these programs, their strengths and weaknesses, ethical practices of using these programs, and how they can avoid their misuse.

For further reading on the risks and concerns surrounding AI in higher education, check out the articles below:

Overall concerns and ethics

• AI Platforms like ChatGPT Are Easy to Use but Also Potentially Dangerous (Scientific American)

• Blueprint for an AI Bill of Rights (The White House)

• Ethical and social risks of harm from Language Models (DeepMind)

• Statement on AI Risk (Center for AI Safety)

• The uselessness of AI ethics (AI and Ethics article)

Bias & Misinformation

• AI Misinformation: How It Works and Ways to Spot It (CNET)

• A developer built a 'propaganda machine' using OpenAI tech to highlight the dangers of mass-produced AI disinformation (Business Insider)

• Exclusive: GPT-4 readily spouts misinformation, study finds (Axios)

• Forecasting potential misuses of language models for disinformation campaigns and how to reduce risk (OpenAI)

• Reflections before the storm: the AI reproduction of biased imagery in global health visuals (The Lancet Global Health)

• Webinar: How to fact-check in the AI Age (News Literacy Project)

Privacy

• Center for AI and Digital Policy FTC Complaint

• How to Delete Your Data from ChatGPT (Wired)

• How Zoom’s terms of service and practices apply to AI features (Zoom blog)

• Is ChatGPT's use of people's data even legal? (Avast blog post)

Plagiarism and Academic Integrity

• Cheating Fears Over Chatbots Were Overblown, New Research Suggests (New York Times)

• Paper exams, chatbot bans: Colleges seek to ‘ChatGPT-proof’ assignments (AP News)

• Signs of undeclared ChatGPT use in papers mounting (Retraction Watch)

Copyright and Intellectual Property

• AI companies would be required to disclose copyrighted training data under new bill (The Verge)

• Artificial intelligence and copyright (World Intellectual Property Organization magazine)

• The New York Times prohibits using its content to train AI models (The Verge)

• Plagiarism Engine: Google’s Content-Swiping AI Could Break the Internet

Another issue that has arisen surrounding generative AI programs relates to copyright and intellectual property.  Since the production of generative AI programs is just being introduced to academics, there is a lack of government policies and protections surrounding confidential information.  There is also confusion surrounding the ownership of AI-generated material since most of this material has minimal to no human input.  To learn more about copyright and intellectual property laws surrounding AI, check out the articles below:

• Generative Artificial Intelligence and Copyright Law (Congressional Research Service)
• Copyrightable Authorship: What Can Be Registered (U.S. Copyright Office)
• Copyright Registration Guidance: Works Containing Material Generated by Artificial Intelligence (U.S. Copyright Office)
• When AI is Writing, Who is the Author?  (Inside Higher Ed)
• OpenAI Sharing & Publication Policy (OpenAI)
• Standards around generative AI (The Associated Press)

Bailey, J. (2023). AI in Education: The leap into a new era of machine intelligence carries risks and challenges, but also plenty of promise. Education Next, 23(4), 28-35.                                                                       

Mogos, A., Ahmad, Y. & Culver, M. (2023, August 11). The Promises and Perils of Generative AI in Education: TFA’s Evolving Perspective. https://www.teachforamerica.org/one-day/ideas-and-solutions/the-promises-and-perils-of-generative-ai-in-education-tfas-evolving

Stimpson, J. (2023). ChatGPT and Generative Artificial Intelligence: Concerns about AI. Massachusetts Library System. https://guides.masslibsystem.org/ai/concerns 

Research, H. (2023, September). Benefits, Challenges, and Sample Use Cases of Artificial Intelligence in Higher education. https://www.insidehighered.com/sites/default/files/2023-10/Benefits%2C%20Challenges%2C%20and%20Sample%20Use%20Cases%20of%20AI%20in%20Higher%20Education.pdf