Artificial Intelligence

Generative AI is great for brainstorming, summarizing, and tutoring, but it’s not a substitute for verified academic sources.
Search engines help locate a wide range of content but require careful evaluation of credibility.
Library databases are the gold standard for scholarly research, offering peer-reviewed, citable sources.

Remember: If you are not an expert (or at least somewhat knowledgeable) on a topic, you cannot expect or trust AI to be

AI Integrations in Databases

Across the board, academic database providers are weaving AI into their services in 2025, but each in a way that aligns with their platform’s purpose:

• Content aggregators like EBSCO and ProQuest use AI to simplify search and help users digest articles (summaries, concept explanations).
• Digital libraries like JSTOR use AI to enhance reading comprehension and exploration without replacing the scholarly material.
• Major publishers like Elsevier deploy powerful generative AI to let researchers query vast collections and speed up literature reviews (with rigorous source attribution to maintain trust).
• Educational platforms like Gale focus on AI that aids discovery and analytical skills (topic finding, text analysis), keeping the researcher in the loop.

In all cases, the goal is to save researchers time and effort while preserving accuracy and scholarly integrity. Users should always remain critical thinkers – the AI tools provide assistance, not absolute answers – but when used wisely, these features can significantly enhance the research experience, from quickly screening hundreds of papers to engaging interactively with a single book chapter. Each database’s approach to AI offers unique strengths, and together they point toward a future where scholarly research is more efficient and richly supported by “intelligent” tools.

AI integration in academic databases falls into a few key categories:

• AI-Assisted Search: Natural language query handling, semantic search improvements, and AI-driven query expansion.
• Automatic Summarization: On-demand summaries or key point extractions from articles to help researchers quickly gauge relevance.
• Conversational AI/Chatbots: Interactive tools that allow users to ask questions about a document or topic and get answers with references.
• Recommendation & Analysis Tools: Features that suggest related content, compare findings across papers, or perform text analysis (topic modeling, etc.).

Content was developed using responses generated with Copilot Researcher using two prompts: "use this list of databases https://libguides.xavier.edu/az.php to find which ones have added language to their sites and contracts about artificial intelligence" and "what AI features each of these databases now include?"

Privacy Tips

• Protect confidential data
  • Do not enter personal or confidential data into publicly-available Generative AI tools. Information shared with Generative AI tools using default settings is not private and could expose proprietary or sensitive information to unauthorized parties.
• Review content before publication
  • AI-generated content can be inaccurate, misleading, or entirely fabricated (sometimes called “hallucinations”) or may contain copyrighted material. You are responsible for any content that you publish that includes AI-generated material.
• Adhere to existing academic policy
  • Review Xavier’s student and faculty handbooks and policies. Xavier will be developing and updating their policies as we better understand the implications of using Generative AI tools. In the meantime, faculty should be clear with students they’re teaching and advising about their policies on permitted uses, if any, of Generative AI in classes and on academic work.Students are also encouraged to ask their instructors for clarification about these policies as needed.
• Be alert for phishing
  • Generative AI has made it easier for malicious actors to create sophisticated phishing emails and “deepfakes” (i.e., video or audio intended to convincingly mimic a person’s voice or physical appearance without their consent) at a far greater scale.
• Disable model improvement or data sharing
  • Location varies model to model but in your profile settings, check under Privacy, Activity, or Data Controls to opt out.

Before you jump into using GenAI for research, try these warm-up exercises to get a feel for the responses...and to see if you are smarter than a GPT!

• Ask a series of questions about a topic you know a lot about (what are you an “expert” in) (does not need to be academic)
  • Are the answers accurate?
  • Does it provide links to resources about the topic?
  • What happens when you correct it?
• Ask it to write a story about two fictional characters from different worlds (IPs) meeting in one of the character’s worlds. Then ask it to write the story set in the other character’s world. 
  • Was the depiction of the characters accurate?
  • Was the writing entertaining?
  • Would you want to read/watch more of the story it produced?

Generative AI Research Tools

• ResearchRabbit
  • Search for papers and authors, monitor new literature, visualize research landscapes, and collaborate with colleagues.” -Mission statement
  • Free and search powered by PubMed and Semantic Scholar
• Semantic Scholar
  • Free, AI-driven search and discovery tools, and open resources for the global research community.
• Keenious
  • Analyzes inputs (text, URL, pdf) to find relevant scholarly articles and related topics
  • Free for core features with limited results and analysis, $10/month for expanded results and unlimited analysis.

Productivity AI Tools

• Goblin Tools
  • A collection of small, simple, single-task tools, mostly designed to help neurodivergent people with tasks they find overwhelming or difficult.
  • Most tools will use AI technologies in the back-end to achieve their goals. Currently this includes OpenAI's models. As the tools and backend improve, the intent is to include ethical alternatives.
• Perplexity
  • Perplexity AI is an innovative AI-powered search engine that combines the capabilities of large language models with real-time web searching to provide comprehensive, up-to-date answers to user queries. 
  • Perplexity is available as a web app, mobile app, and Chrome extension. Standard plan: free forever. Pro plan: $20 per month or $200 per year.
• NotebookLM
  • AI research tool and thinking partner powered by Google Gemini. It can analyze your sources, turn complexity into clarity, and transform your content into a variety of digital mediums. NotebookLM will search only through the sources you provide it. Learn more about the 2025 updates.

The Library does not endorse any specific AI technologies, and encourages users to be cautious about sharing personal information when using AI tools. 

4 P’s of Prompting

1. Priming–Give it context.
2. Purpose–Give it a task.
3. Persona-Give it a descriptive profile.
4. Prompt Tuning –Iterate, separate

For example:

You are a Google Cloud program manager (1). Draft an executive summary email to (2) [persona] based on [details about relevant program docs] (3). Limit to bullet points (4)

5 Prompting Techniques

1. Choose the best available model
   • Premium (paid) and newer models have improved accuracy, reasoning, reliability, and speed. They follow instructions better and make fewer error.
2. Decomposing the Problem
   • And the end of your prompt, try adding "before answering, list the sub-problems
     you need to solve and ask me questions to help us get this done”
3. Self-Criticism
   • After the output is generated, follow-up with a prompts like "look at your own work and find the weak spots" or "what is incorrect with your output?" or "what is wrong with your output?"
4. Context Engineering
   • The more relevant information you give, the better the AI understands your task. Provide background up front.
   • Share your goals, constraints, audience, or examples so the model can “think” in the right frame.
   • words perfect in prompt < context and detail to give background information
5. Ensembling
   • Use multiple models to leverage complementary strengths by designing workflows: one model drafts, another
     critiques, another polishes
   • Reduce blind spots and errors. Combining models helps balance biases, catch mistakes, and surface new angles.
   • Boosts accuracy, creativity, and reliability.

Prompting Tips

• Use natural language.
  • Write as if you’re speaking to another person.
  • Express complete thoughts in full sentences.
• Be specific and iterate.
  • Tell the program what you need it to do (summarize, write, change the tone, create).
  • Provide as much context as possible.
• Be concise and avoid complexity.
  • State your request in brief —but specific —language. Avoid jargon.
  • Most successful prompts average 21 words, but most people post only 9-word prompts.
• Make it a conversation.
  • Fine-tune your prompts if the results don’t meet your expectations or if you believe there’s room for improvement.
  • Use follow-up prompts and an iterative process of review and refinement to yield better results

Prompting Resources

• Write AI Prompts
• AI for Education Prompt Library
• Prompt Engineering Guide Examples
• Gemini for Google Workspace Prompting Guide 101
• Learn About Microsoft Copilot Prompts
• Generative AI Prompt Book (Feb 2024)