What do we mean when we say AI?
The term “AI” is everywhere. It’s hard to know what anyone means when they use it.
Most of the AI in our daily tools isn’t new. The technology behind spam filters, autocorrect, social media algorithms, your Netflix queue, and search engine results has been part of our digital lives for years. We’re already using it.
What’s new is generative AI: tools like ChatGPT, Claude, Gemini, and Microsoft Copilot that can write a news release, generate a photorealistic image, or produce a video from a text prompt alone. These are powered by Large Language Models (LLMs) and trained on large, generalized, publicly available data to generate language that sounds fluent and plausible. (This is different from AI that is embedded in tools like Adobe, ClickUp, or Sprout, which are trained for specific tasks using specific data.)
LLMs can produce confident-sounding content that is wrong. Most publicly available LLMs have no judgment, no institutional knowledge, and no inherent stake in IU’s reputation. You have all three.
Using AI as a co-intelligence—a collaborator that supports and complements human thinking—is encouraged, with your human judgment and IU’s values guiding every step. These suggestions focus on generative AI, the technology most familiar to us as marketing and communications professionals, rather than agentic AI, which automates task completion in ways that raise a different set of considerations.
Our approach
We believe AI should amplify human potential, not diminish it. Everything we create is anchored by people and their judgment. Our institutional knowledge, our relationships, our understanding of what makes IU’s story worth telling begins with us.
The university is building a track record of responsible AI implementation across teaching, research, and service. University Communications and Marketing drives the story. Understanding how AI is reshaping the way audiences find, filter, and connect with content isn’t optional, it’s core to our work.
Used thoughtfully, AI has real potential to transform how IU fulfills its mission: deeper student success, stronger research, more meaningful public service. We’re committed to being on that frontier. Using AI innovatively and using it responsibly aren’t in tension. The goal is both.
New to generative AI or want to build a stronger foundation?
IU’s free, self-paced GenAI 101 course is a great place to start. Available to IU community members and the general public, it offers practical skills, no prior experience needed.
New learning pathways for specific industries and professions are in the works!
Prospective and current students have AI radar
Gen Z (born 1997–2012) and Gen Alpha (born 2013 and after) have grown up with generative AI.
They can detect content that feels produced rather than written, and images that feel generated rather than real. Authenticity connects with the people we most want to reach.
Where AI earns its place
Before you create
- Suggesting a structure when your ideas are there but the shape isn’t
- Generating a dozen headline options when you’re stuck
- Brainstorming questions before an interview
- Brainstorming photo locations or compositions
- Pointing you toward the most relevant sources faster than a search would
After you create
- Finding gaps in an argument you’ve already made
- Reframing a complex idea for a different audience
- Helping you turn a dense paragraph into something scannable
- Checking whether your tone matches your intent
- Seeing your draft differently when it isn’t working
None of these replace your thinking. All of them can make it sharper.
How we work with AI
Start here
Practices and habits worth building.
Use judgment
From brainstorming to final edits, these are the places where your judgment makes all the difference.
| The idea | The action |
|---|---|
| AI can expand what you’re already thinking. | Use it to generate alternate headlines, angles, or approaches, then decide what’s worth pursuing. |
| Quantity isn’t quality. | Ask for ten options knowing you’ll use none of them verbatim. You’re looking for the spark, not the answer. Bring your own structure, voice, and ideas first, then ask AI to pressure-test or reframe. |
| AI can help you ask better questions. | Use it to explore a topic or develop interview questions, then verify everything independently. |
| A second read is always useful. | Ask AI to check tone, clarity, or consistency, then decide what feedback actually applies. Use AI for storyboarding, composition planning, and early mockups, not the final public-facing assets. |
| AI reflects the biases in its training. | Read what it produces with an editorial eye on which perceptions it centers or marginalizes. |
Hold the line
Practices that protect our credibility.
| The idea | The action |
|---|---|
| Fully AI-generated content isn’t ours. | Don’t publish any content that hasn’t been vetted and edited by a human. |
| AI images aren’t authentic. | IU photography is our strongest visual asset. If you need stock, choose images that feel genuine. See visual content considerations for additional guidance. |
| Prompts are public spaces. | Never enter student data, embargoed information, or anything protected by FERPA or HIPAA. Check your tool’s data settings to ensure your inputs aren't being used to train models. (See IU’s Management of Institutional Data policy.) |
| Attributed words carry weight. | Every published quote needs a real person’s approval. |
| AI doesn’t fact-check itself. | Verify everything independently. What’s plausible isn’t necessarily accurate. |
Additional considerations
The AI that is embedded in the tools we use, such as Microsoft Teams, Zoom, Slack, ClickUp, and Adobe, works differently than public LLMs like ChatGPT. Knowing the difference helps you use each one responsibly.
When in doubt, check the tool’s privacy settings or get help from UITS.
A few additional considerations for writers, storytellers, strategists, marketers, and anyone creating content for web, social, email, or campaigns.
- When sharing AI-assisted outlines or briefs with colleagues, flag that AI helped shape the structure so your team can build on your thinking rather than the model’s.
- When asking AI to help with research or background, prompt it to identify sources for factual claims, then verify each one yourself.
- Reformatting existing content for different platforms (web to email, longform to social) is one of the most practical everyday uses of AI.
- Strong human-written content remains one of the most effective tools for SEO/AIO visibility. Authentic, specific writing is rewarded by both audiences and algorithms.
A few additional considerations for designers, photographers, videographers, and anyone working with visual assets.
- Photo-realistic AI-generated images carry legal risk around rights of publicity and likeness — and your audiences will spot them. A prompt as simple as 'IU fan at a tailgate' could produce an image resembling a real person who hasn't consented to appear in university materials. Inauthentic imagery erodes brand trust. Use real people.
- Anything you use generative AI to create, including data visualizations and graphics, is unsettled legal territory as it relates to intellectual property law.
- Before reaching for an AI tool, check IU’s branded design templates for a foundation that’s already aligned with IU’s visual identity.
- Never use AI to recreate or approximate IU brand marks. AI will distort, rearrange, or alter official marks in ways that undermine consistency and ownership. Always use official, unaltered IU brand assets.
This is a starting point
AI is evolving quickly, and so is our thinking. These practices reflect where we are now: grounded in our values and open to what comes next.
A note on how this page was made
These best practices were developed by IU’s marketing and communications community. In drafting this page, an AI assistant was used to help organize structure, explore language, and refine ideas. Every recommendation, decision, and judgment call remained human.