AI Tools & Google Education

Explanation of this List

This list is intended to give you an overview of what we are currently able to learn about the following AI tools based on the publicly-available information. We will attempt to give you information on recommended use cases, how datasets are compiled, biases that may be present in the training datasets, and so on.

• Don't see a tool you want to know about in this list? Check out Georgetown's table of AI Resources, or contact digital-learning-studio@willamette.edu and we will add to this list. 
• More general information about LLMs is available on the General Resources page of the Willamette Library Artificial Intelligence (A.I. Resources) guide.
• Please note that you should always fact-check any information generated by a LLM.

Table of Contents:

• NotebookLM
• Gemini
• Adobe Firefly
• ChatGPT
• Claude
• Midjourney
• Napkin AI
• Perplexity

NotebookLM

Suggested uses / what it does

• Collects information from different types of sources and creates custom notebooks.  It is good at extracting information from a specific source and citing where it is found.

Training data

• Works with uploaded sources (pdfs, web pages, YouTube, mp3 files, Google docs

Biases

• N/A, only those inherent in uploaded sources

Ethical issues

• Even with Google's privacy assurances, there are questions about data storage, misuse especially when dealing with confidential or personal information.

Is it free?

• Free with a Google account (we recommend using your Willamette Chrome login)

More information

• NotebookLM does not use your personal data, including your source uploads, queries, and the responses from the model for training.

Gemini

Suggested uses / what it does

• The Gemini LLM allows you to query data you upload in the form of prompts, files, images, alongside materials available on the internet. 

Training data

• Gemini is trained on publicly available code (read: much of the public internet) and Google Cloud materials

Biases

• Reflective of the current biases present on the internet, and those in the training dataset materials.

Ethical issues

• Even with Google's privacy assurances, there are questions about data storage, misuse especially when dealing with confidential or personal information.

Is it free?

• Free with a Google account (recommend using Willamette Chrome login)

More information

• Google collects your chats, information you share with Gemini, information about the location of your device, IP address, and home or work addresses in your Google account. This data is used “to provide, improve, develop, and personalize Google products and services and machine-learning technologies, including Google’s enterprise products such as Google Cloud.” – Gemini Apps Privacy Hub page
• Human reviewers DO “read, annotate, and process your Gemini Apps conversations” although steps are taken to protect privacy. As such you should NEVER enter confidential or personal data in your prompts. Learn more on the Gemini Apps Privacy Hub page.

Adobe Firefly

Suggested uses / what it does

• create video from still images
• create visual effects
• create audio effects
• audio and video editing
• translation features
• collaborative boards

Training data

• “Feel confident creating with Firefly knowing it’s safe for commercial use. Adobe does not train Firefly on customer content or content mined from the web. We train our models only on content where we have permission or rights to prevent it from creating content that infringes copyright or intellectual property rights.” – Adobe's AI ethics statement

Biases

• N/A

Ethical issues

• According to Adobe's AI ethics statement, Firefly is not trained on customer content, and Adobe compensates creators for use of their content to train Firefly. Adobe does not claim ownership of anything you create in Firefly.
• It should also be noted that Adobe has partnered with Google, OpenAI, and other AI companies to use their tools in Firefly – see this list on their website – so ethical concerns with those companies may come into play.

Is it free?

• Willamette has access to Firefly through our Creative Cloud license. Use the full program on one of WU’s Adobe-licensed computers (in the Digital Learning Studio, art building, etc), or get a license for Adobe Express from WITS and use a limited version of Firefly with that app.

More information

• While Adobe says that content created with Firefly is safe for commercial use, this does not apply if you add copyrighted material into your projects, so be vigilant in checking the copyrights of materials that you bring to the table. See Firefly’s GenAI user guidelines on their website.
• It should also be noted that Adobe has partnered with Google, OpenAI, and other AI companies to use their tools in Firefly – see this list on their website.
• As of March 2025, the US court system has ruled that AI-generated content cannot be copyrighted, human-authorship is necessary for copyright. (See Thaler v. Perlmutter; summarized on Stanford University Library's Copyright & Fair Use page). 

ChatGPT

Suggested uses / what it does

• Generalized LLM. ChatGPT searches the internet to respond to queries.

Training data

• OpenAI claims that its models are trained on publicly available information on the internet, information from third-party partners, and information that users provide or generate.
• There are several lawsuits against OpenAI regarding copyright infringement.

Biases

• Reflective of the current biases present on the internet, and those in the training dataset materials.

Ethical issues

• There are several lawsuits against OpenAI regarding copyright infringement.

Is it free?

• Free and paid options.

More Information

• While OpenAI claims it does not intentionally collect personal information, its LLMs train on data from and about people. You should never put personal data into prompts.
• OpenAI collects personal data if you open an account with them, including: contact information, account credentials, birthday, payment information, content you upload, contents of any information you send them through email, social media, or surveys, your log and usage data, device information, and location information. 
• OpenAI will share your personal data with government authorities and other third parties.
• Learn more in OpenAI's privacy policy

Claude

Suggested uses / what it does

• The Claude LLM allows you to query data you upload in the form of prompts, files, images, alongside materials available on the internet. Suggested uses: coding support, data querying. 

Training data

• Anthropic does not provide specific information on where training data is sourced but does mention using internet sources. It appears they also collect information from users to train their models. 

Biases

• Reflective of the current biases present on the internet. 

Ethical issues

• Anthropic has allegedly pirated copyrighted books and music to train their LLM. They are currently being sued for copyright infringement. This is especially damning due to the company claiming to be an “ethical ai company.”

Is it free?

• Free & paid options available.

Midjourney

Suggested uses / what it does

• Text to image creation LLM, image to video creation.

Training data

• Midjourney does not cite its training sources, but it is being sued by Disney, Universal, and groups of artists. 

Biases

• Dependent upon the art used to train the LLM. 

Ethical issues

• Trained on copyrighted images, and on images without permission of creators.

Is it free?

• No, there are different levels of paid plans.

More Information

• Currently being sued for copyright infringement by Disney and Universal, artists have also filed a class action lawsuit.

Napkin AI

Suggested uses / what it does

• Text to image infographic creation.

Training data

• Napkin does not mention its training sources.

Biases

• N/A

Ethical issues

• Napkin AI runs on an LLM driven by OpenAI's GPT-4o, so ethical concerns related to OpenAI apply.

Is it free?

• Free & paid options are available.

More information

• Napkin AI collects data you submit to the visual AI, but you can opt out of data collection.
• You own the content you create.
• You may not resell generic output that Napkin generates, such as icons, illustrations, or collections. 

Perplexity

Suggested uses / what it does

• Perplexity is a web search engine that uses current internet content to answer queries. 

Training data

• Perplexity does not provide information about what is contained in their training data but there have been concerns about privacy and copyright infringement.

Biases

• Reflective of the current biases present on the internet. 

Ethical issues

• The parent company has gotten into legal issues due to copyright & trademark infringement and disrespecting Robot Exclusion Protocol on websites. Users have raised concerns about data privacy and information fabrication. 

Is it free?

• Free & paid options.