Let's be honest about something: AI can make it trivially easy to cheat. And if we're building an AI tool for students, we have a responsibility to talk about that directly.
This isn't a post where we pretend the problem doesn't exist. It does. AI has fundamentally changed what academic dishonesty looks like, and any company that builds AI for students without addressing it is being irresponsible. So here's our honest take — what the problem actually is, how we think about it at Nexus Pen, and how we've designed Donna to be a learning partner rather than an answer machine.
The Real Problem With AI and Cheating
The concern isn't really about AI. It's about a deeper question: are students actually learning? A student who copies an essay from ChatGPT isn't learning to write. A student who asks Donna "what's the answer to problem 7" without understanding the method isn't learning math. The harm isn't the technology — it's the shortcut that bypasses the cognitive work that builds real understanding.
The same shortcut existed before AI. Students have always been able to find answers online, copy from classmates, or hire tutors who do the work for them. AI just made it faster and more accessible. That means the solution isn't to ban AI — it's to redesign how we think about learning with AI.
What "Responsible AI Use" Actually Means
We think about responsible AI use in education along three dimensions:
- Transparency — Students should be honest about when and how they used AI. Using Donna to check your understanding of a concept is different from using it to generate an essay submission.
- Engagement — AI should increase your engagement with material, not replace it. If you're asking Donna to explain a concept and then closing your notebook, that's passive consumption. If you're asking Donna to quiz you, challenge your reasoning, or suggest connections to other topics, that's active learning.
- Ownership — At the end of a study session, can you explain what you learned without Donna's help? If not, something went wrong. The goal is always to internalize knowledge, not outsource it.
How We Designed Donna With This in Mind
Donna's School Mode isn't just a feature name — it reflects a specific pedagogical philosophy. When a student asks a direct homework question in School Mode, Donna's default behavior is to ask a clarifying question back, or explain the concept rather than stating the answer. This is deliberate.
We call it the Socratic default. Instead of "The answer is 42," Donna responds with "What do you already know about this type of problem? Let's work through it." It takes slightly longer. It's occasionally frustrating for students who just want the answer. That friction is intentional — because the friction is where learning happens.
We've also built in a concept called answer delay. For certain question types — especially math problem-solving and essay prompts — Donna walks through the reasoning process step by step before arriving at a conclusion. This mirrors how a good tutor works: showing the path, not just the destination.
A Thinking Partner, Not an Answer Machine
The best analogy we've found is this: imagine you have a brilliant friend who has a PhD in every subject. You wouldn't call them up and say "write my essay for me." You'd call them when you're stuck, when you want to test your understanding, when you need someone to push back on your reasoning. That's the relationship we want students to have with Donna.
Good tutors don't give students answers. They ask better questions. They make students feel capable rather than dependent. They celebrate the moments when a student says "oh, I think I get it now" — not because the tutor explained it, but because the student worked through it. Donna is built to create those moments.
The Teacher's Role Hasn't Changed
One thing we want to be clear about: Nexus Pen doesn't replace teachers, and responsible AI use in education isn't something students figure out on their own. Teachers set the norms. They establish expectations for when AI is appropriate and when it isn't. They design assessments that require genuine understanding rather than answer retrieval.
We provide resources for teachers to define clear AI use policies in their classrooms, and our School Plan includes admin-level controls that let teachers configure what Donna can and can't do during specific assignments. A teacher running an in-class essay can put student pens into a restricted mode. A teacher assigning independent research can allow full access.
Our Commitment
We're a team of students ourselves. We know what academic pressure feels like. We've all had moments where the tempting shortcut looked really appealing. We built Nexus Pen because we believe AI can make learning genuinely better — not because we wanted to make cheating easier.
That means we'll keep talking about this honestly. We'll keep refining Donna's behavior. And we'll keep building tools that help students become smarter, not just faster at getting answers.
Because at the end of the day, the goal isn't a good grade. It's understanding the world well enough to do something meaningful with it.