With a name like Grammarly, the software may sound like a simple proofreading service—but did you know it can write outlines and reports? 

Alison Thomas, CAS/MFA ’06, assistant dean for academic integrity, detailed the advantages and limitations of the digital-assistant tool during the Center for Teaching, Research, and Learning’s 35th annual Ann Ferren Conference, January 11–12. Her session, “Personalized AI, Everywhere You Write: Thinking Critically About Grammarly,” prepared faculty attendees to identify use cases for their courses and brief students on the AI’s functionality.

Grammarly emerged in 2009 as a service akin to Microsoft’s spelling and grammar editor. In those first years, Thomas explained, educators recommended it to students with learning disabilities like dyslexia, as well as multilingual and online learners. While “there [are] definitely some assumptions about who uses Grammarly,” Thomas said, “I would correct these . . . [and] say everyone uses [it].” Today, thanks to its souped-up AI properties, Grammarly—which drew more than 30 million users in 2023, up from 1 million just eight years ago—is what Thomas calls a “pervasive” and “universal” writing aid.

“The little green lightbulb,” as Thomas described the tool, can be downloaded to a desktop, imported as a Google Docs or Chrome plugin, or added to an iPhone or Android. Only a click away, the assistant waits to help hone the user’s writing voice. If Grammarly detects a company update, for example, it will offer tips for improving the email, like “Consider including any follow-up actions.” Users can request additional assistance in a chat box. Typing, “Draft a company update” prompts auto-generated text within seconds.

Beyond composing sentences, offering structural advice, and reviewing grammar, the software can write research plans, draft outlines, locate key ideas, and propose counterarguments. It can even brainstorm—one of many capabilities enticing to students, Thomas said.

Understanding the breadth of Grammarly’s capabilities “should allow you to give more specific guidelines about what is and isn’t okay,” she told the 30 or so faculty in attendance. And from the students’ perspective, a well-rounded understanding of the tool allows them to make informed decisions about when powering up Grammarly could be appropriate—or dishonest.

Thomas recommended that Grammarly users keep context top of mind. Students, for example, should ask themselves: Does my reader expect my work to be original? Does using generative AI align with my goals? Will I lose out on a learning opportunity if I lean on machine learning?

When it comes to the ethics of generative AI, most of the battle is knowing how and when to employ it. No one feature is inherently good or bad; its application determines its appropriateness. Thomas raised the brainstorming function as an example. Students could upload an assignment prompt, ask Grammarly for paper topics, then pass one of these off—verbatim—as their own. Or they could import the prompt, review the list to better understand the assignment, then write a wholly original topic—perhaps even stronger than they could’ve generated outright.

While some instructors may maintain neither application is ideal, as Grammarly intervenes to “think” for the student, others may consider the latter scenario a perfectly responsible use of AI. One attendee suggested that all instructors, whether Grammarly skeptics or enthusiasts, could benefit from having students turn in “AI appendices” with assignments. Adding this step would require them to acknowledge their use of AI and reflect on the implications of that decision.

Although there is no consensus on the most ethical way to use it, generative AI is here to stay. Promoting a culture of openness in the classroom and articulating clear guidelines around technology are the best ways for professors to manage expectations in a world swirling with change, Thomas said. Either we learn to live and work with AI, or we get left behind in its pixelated dust.

Faculty seeking additional resources on generative AI can visit the Office for Academic Integrity’s SharePoint site or contact the Center for Teaching, Research, and Learning for support.