Using Generative AI at Dartmouth - Staff Overview

Note: Generative AI tools should be used with care and in compliance with college policy. Please review the Guidelines and Considerations on this page for more information.

What is Generative AI?

Generative AI, short for Generative Artificial Intelligence, is a powerful branch of artificial intelligence that focuses on creating new content, whether it's text, images, audio, or more, often with a human-like touch of creativity. Unlike traditional AI, which typically performs predefined tasks, generative AI models have the ability to produce original and meaningful output without specific programming for every potential result.

Generative AI Tools

There are many Generative AI tools available now, and many more coming out. Some AI tools are separate web-based applications that require signing up for a free or paid account, while others are embedded features in popular applications.

Common AI Tools

Some of the popular Generative AI Tools you may find out in the wild. If you choose to use them, please review their terms of service for information on privacy and cost implications.

OpenAI: ChatGPT, DALL-E
Google: Bard, Duet AI
Microsoft: Bing AI, 365 Copilot

Embedded AI

Many of the software solutions you use already have started to embed generative AI features into their features.

You may also find that these features require additional payment or licensing to use. Please consult with ITC for further information on activating these features.

Dartmouth ITC does not currently provide any Generative AI tools outside of software applications with embedded features.
If you need to procure a Generative AI product, please fill out an ITC Rapid Review

How Can AI Assist in Your Work

Generative AI can be a valuable tool to improve efficiency, streamline processes, and enhance decision-making. Remember to consider data privacy and security when implementing generative AI in your work, especially when dealing with sensitive information. Generative AI can provide a great jumping off point to start your projects, boost productivity and improve the quality of administrative work when integrated thoughtfully into your daily tasks.

Dartmouth supports responsible experimentation with generative AI tools, but there are important considerations to keep in mind when using these tools. All Dartmouth community members must adhere to Dartmouth’s existing policies, including the Confidentiality of Business Information Policy, Dartmouth Information Security Policy, and Nondiscrimination and Anti-Harassment Policy and Resolution Procedures.

Here are some ways AI can be used:

  • Document Generation
    Use generative AI to automatically generate reports, memos, emails, or other written documents. Tools like GPT-3 can assist in composing professional and coherent text.
  • Data Analysis
    Employ AI-driven data analysis tools to process large datasets quickly and extract valuable insights. AI can help identify trends, anomalies, and patterns in data, aiding decision-making.
  • Email Management
    Use AI-based email classification and prioritization to automatically sort and categorize emails. This can help you focus on essential messages and reduce email overload.
  • Scheduling
    AI-powered scheduling assistants can analyze your calendar and suggest optimal meeting times, reducing scheduling conflicts and time spent on coordination.
  • Chatbots for Support
    Integrate AI-driven chatbots into your organization's website or messaging platforms to provide instant support and answer common questions from employees or clients.
  • Proofreading and Editing
    AI tools can assist in proofreading and editing documents for grammar, spelling, and style errors, ensuring that your written communications are polished and error-free.
  • Expense Management
    Automate expense reporting with AI tools that can extract data from receipts and categorize expenses, making the process more efficient and accurate.
  • Task Automation
    Implement AI-driven task automation for routine administrative tasks such as data entry, file organization, or form filling, freeing up your time for more strategic work.
  • Language Translation
    Utilize AI-powered translation tools to communicate with colleagues or clients in different languages, breaking down language barriers.
  • Document Summarization
    AI models can summarize lengthy documents or meeting transcripts, saving you time when reviewing and extracting key information.
  • Virtual Meetings
    AI can enhance virtual meetings by providing real-time language translation, transcription, and even automating meeting minutes.
  • Knowledge Base Creation
    Use generative AI to create and update internal knowledge bases, FAQs, and training materials, ensuring that information is always up-to-date.

Data Protection Requirements

Information shared with generative AI tools using default settings may not be private and could expose proprietary or sensitive information to unauthorized parties. ITC evaluates Generative AI (GAI) tools to ensure compatibility with Dartmouth's information security policies. Approved solutions must adhere to relevant security policies. Users are responsible for maintaining compliance when utilizing these approved tools.
 
Data at Dartmouth is categorized into three tiers, based on the likelihood of being targeted by external parties and the potential impact on the organization if exposed. This classification determines the required level of data protection and helps in selecting appropriate GAI tools.

Note:

  • Only approved GAI tools are authorized for handling of DISC 2+ data.
  • For DISC 2 data, contact ITC if your preferred tool is not listed.
  • For DISC 3 data, consult with ITC to ensure adequate security measures are in place.

Data Classification Process:

  1. Data Inventory: Catalog all data to be inputted into a GAI tool, including both analysis data and prompts for generating work products.
  2. Output Classification: Continue to catalog all data outputs from the GAI tool, like data tables or email messages.
  3. Classification Steps: For each cataloged item:
    • Likelihood: Rate the chance of a security breach or incident impacting the data (1: least likely, 3: most likely)
    • Impact: Evaluate the severity of the incident, considering financial, legal, reputational, operational and data integrity consequences (1: minimal impact, 3: major impact)
    • Classification: Add the numbers for likelihood and impact, divide by two and round down. The result will be a number from 1-3 and is this data’s DISC classification.  
  4. Overall Classification: Determine the highest classification level among all data items. Use this to select a suitable GAI tool.

Other Considerations

Responsibility and Accuracy

AI-generated content can be inaccurate, incomplete, misleading, entirely fabricated, or may contain copyrighted material due to its training bias. Review your AI-generated content before publication or use in institutional decision-making, ensuring alignment with Dartmouth’s values and mission. Additionally, consider the appropriateness of Generative AI for the intended use and prioritize privacy protection for all involved. 

You are responsible for any content that you produce or publish that includes AI-generated material. This responsibility extends to assessing potential risks to individuals or the institution. Furthermore, be mindful of the ability to explain and justify Generative AI's decision-making process, maintaining transparency and accountability.


Adhere to Current Policies On Academic Integrity

For those working with faculty or students, please review the relevant School’s student and faculty handbooks and policies (i.e., faculty handbooks and policies of the College, Geisel, Guarini, Tuck, and/or Thayer). It is important to remain compliant with specific rules, laws, regulations, including institutional policies and data use agreements. Ensure that the use of generative AI tools aligns with Dartmouth’s values and mission, and conduct a thorough assessment of the appropriateness of such tools for the intended use.

We expect that Schools will be developing and updating their policies as we better understand the implications of using generative AI tools. You can find information about teaching with generative AI on DCAL’s website. In this evolving landscape, consider the privacy protection of all involved parties and remain vigilant about potential risks to individuals or the institution.


Be alert for AI-enabled phishing

Generative AI has made it easier for malicious actors to create sophisticated scams at a far grander scale. Continue to follow security best practices and report suspicious messages to phishing@dartmouth.edu.


Connect with Dartmouth ITC before procuring generative AI tools

Dartmouth is working to ensure that tools procured on behalf of Dartmouth have the appropriate privacy and security protections and provide the best use of Dartmouth’s funds.

  • If you have procured or are considering procuring generative AI tools or have questions, Please fill out a ITC Rapid Review
  • All generative AI tools must be assessed for risk by Dartmouth’s Information Security and Data Protection team, please fill out a ITC Rapid Review

Coming soon

Pre-Approved Tools

Tool Classification
Example 1 Lorem ipsum dolor sit amet, consectetur adipiscing elit.
Example 2 Quisque vel ante facilisis, hendrerit felis ullamcorper, porttitor nibh.

How to Get Help

If you have questions or need help, please reach out to Client Technology Consulting.

Submit Ticket

How to Procure

If you are looking to purchase or procure Generative AI tools, please work with Information, Technology and Consulting to ensure your software meets all the security and technology requirements of the institution.

ITC Rapid Review

Details

Article ID: 155837
Created
Mon 12/4/23 3:24 PM
Modified
Mon 1/22/24 12:12 PM