Artificial Intelligence (AI)

What is Artificial Intelligence (AI)?

  • Artificial Intelligence (AI) is the phrase used to refer to computer systems that mimic or exceed human thinking based on experience. AI systems rely on algorithms and models that can analyze large data sets to identify patterns for predictions or decision-making.

What is generative AI?

  • Generative AI is a form of AI used to create new content, images, text, videos, music, etc., based on user inputs. Generative AI uses large data sets and algorithms to analyze patterns and rules to “learn” behavior and characteristics of training data sets, and it uses that “learned” information to produce materials that can be superficially convincing, but not always persuasive or accurate.

What is ChatGPT?

  • ChatGPT is a natural language AI that allows the user to ask questions conversationally. It can help compose emails, essays, and even programming code.

What are AI image generators?

  • AI image generators create realistic images and art from text descriptions or prompts. Two of these generators are DALL·E 2 and Midjourney.

What are AI hallucinations?

  • Hallucinations are when AI generates false information, perceptions of something that is not actually present or factual. Because AI is trained on large data sets to learn patterns, they may create new text, not necessarily based on facts.

What is AI bias?

  • AI bias occurs when algorithms return systematically biased information based on incorrect assumptions. This can happen due to biases in the data used to train the systems or algorithms.

What information should I not feed a generative AI tool?

  • Data classified as private or restricted should not be entered into generative AI tools, including non-public research data, per SBHE 1202.3 Data Privacy Policy. Do not disclose confidential, sensitive, or personally identifiable information when using these tools. Do not disclose intellectual property that is not safeguarded. This caution extends not only to sensitive data generated and used as part of a research project but also to protected student data and student information.

How can I subscribe to a generative AI tool?

  • Contact IT Central prior to purchasing any software applications that will be used on Minot State computers.

As we start another academic year, please consider some initial guidelines similar to those at other institutions across the nation (e.g., see the statements from Harvard University, University of Delaware, Washington University in St. Louis). The following guidelines were developed by UND and adapted to fit Minot State with their permission.

Understand Your Responsibility – Content created by generative AI can include factual errors or inaccuracies, fabrications, bias, or other unreliable information. It is your responsibility to ensure the accuracy of the information reported in your work. Review all material produced for accuracy, violations of copyright protections, and plagiarism. Document and be transparent about all use of generative AI—such clarity in citation and attribution is a critical aspect of any research product that uses generative AI. Be sure to comply with academic and research integrity policies: review the Code of Student Life and Faculty Handbook.  In the classroom, be clear about expectations for student use of AI in their coursework.

Guard Confidential Data – Data classified as private or restricted should not be entered into generative AI tools, including non-public research data, per SBHE 1202.3 Data Privacy Policy. When using these tools, do not disclose confidential, sensitive, or personally identifiable information. Do not disclose intellectual property that is not safeguarded. This caution extends not only to sensitive data generated and used as part of a research project but also to protected student data and student information.

Personal Security Starts with You – Be extra vigilant about potential phishing attacks. generative AI is rapidly changing the phishing landscape, and AI has made it possible for bad actors to deploy newer, more sophisticated phishing attacks and other attacks on your personal data and identifying information. Report any questionable emails to Information Technology Central (ITC) by clicking the Phish Notify button in Outlook/Office 365 or submitting a helpdesk ticket.

Reach out to ITC Before Procuring Generative AI Tools – Please contact ITC prior to purchasing software applications. If you have already purchased a generative AI tool, please let ITC know.

What can I, as faculty, do to support academic integrity?

  • Provide clear expectations and rationales regarding academic integrity. Be sure to clarify for students your expectations regarding using any generative AI tools or applications.  State these expectations clearly on your course syllabus and in any assignment prompts, and be sure to explain the consequences for the students if your expectations regarding the use of generative AI tools are not met. 
    • Discuss with them at the beginning of the course and frequently thereafter.
    • Place clear statements in the course syllabus and on Blackboard.
    • Delineate whether using an automated tool such as ChatGPT is considered plagiarism.
    • Note that websites that purport to detect the use of AI are flawed and have high occurrences of false positives and negatives. Any use of these should be transparent and students must consent to faculty submitting their work.
    • Discuss with your students the challenges and opportunities that AI and automation present within your academic discipline and the subject of your courses. Acknowledge that other disciplines, courses, and faculty may have different expectations and understandings of appropriate use.
  • Be transparent about your own uses of AI. This models good practice and contributes to a broader conversation about the potential uses, benefits, and challenges.
  • Report academic integrity concerns by raising an academic honesty flag in Starfish.
  • We encourage policy discussions within your department about transparency and academic integrity around institutional, faculty, and student use of these tools. Greater understanding and clarity will benefit students as well as programs.

Can I use generative AI for course delivery and assessment?

  • AI tools might be useful in teaching and assessment and could drive new course delivery methods. Minot State will continue to have conversations and provide opportunities to learn more about AI and teaching. We encourage you to have conversations within your department about appropriate use in teaching and learning.
  • Be wary of claims by third-party vendors and look for provable results before adopting such tools in your courses.

Can I use generative AI in my research and scholarship?

  • Your work is considered to be your own. Document any use of generative AI in constructing your work, research, and manuscripts. Similarly, if quantitative or computing tools are used in your research, document those uses.
  • Be sure to check your disciplinary organizations and the publishing guidelines of your target journals to fully understand their position on the use of generative AI and large language models in the construction of manuscripts.
  • Similarly, be aware of current guidelines by granting agencies on using generative AI and large language models.
  • Generative AI and large language models are not designed to establish proof or provide accurate facts. It is the responsibility of the human author to ensure the accuracy of what’s reported in your work and to disclose the role of AI in your work.
  • Submitting your work to an AI tool may make your work widely available in that tool’s training set or database. This might put your results into the public domain before you are ready to share them and ahead of peer review, a critical component of the research process.

Does using AI on an academic assignment violate MSU’s academic honesty policy?

  • It depends.
    • In some courses, use of AI may be viewed as plagiarism since attribution of sources is not a feature of generative AI.
    • For some types of assignments, faculty may specify in the syllabus or the assignment directions how and to what extent AI is permitted. Because this will vary by discipline and course level, students should verify this for every assignment and course.
    • In the absence of any statement on an assignment or syllabus, it would be prudent to assume use of AI without attribution is equivalent to “claiming the words of others as your own,” which violates MSU’s expectations for academic honesty.
    • In summary, unless AI is allowed in a class, its use to write responses or generate images or other content may be considered plagiarism.

Can I use AI for my course assignments?

  • Different fields, courses, and instructors will have different policies and guidelines for how AI can or cannot be used. It is important not to make assumptions about what is allowed and to ask for clarification when needed.
  • When submitting your work for credit, it is assumed to be your original work. The use of other resources, including generative AI models, like ChatGPT or Bard or DALL·E 2 and Midjourney, must be documented.
  • Generative AI and large language models are not designed to establish proof or provide accurate facts. It is your responsibility to ensure the accuracy of what you submit.
  • When in doubt about what’s allowed in a given course, clarify it with your professor/instructor.

If you have additional questions about AI, please submit them here.

 
Special thank you extended to the University of North Dakota who provided much of this content with permission to share it and adapt it to fit Minot State's campus community.