Cubby's Ethical AI Manifesto
Our north star: deepen understanding, accelerate practice, and uphold academic integrity.
In 1995, Steve Jobs argued that everyone should go to law school — not to become lawyers, but because "it teaches you how to think." He understood what legal educators have long known: Law school isn't about memorizing cases or mastering doctrine. It's a three-year exercise transforming how you approach problems, construct arguments, and think through complexity. This is why it's still taught using century-old cases that rarely appear in modern practice. The content is just the medium. The real product is a trained legal mind. This is the context in which we need to discuss AI in higher education.
The current conversation around AI in law school is dominated by fears that students will outsource their thinking and that technology will erode academic integrity and enable sophisticated cheating. These fears are not unfounded. Generic AI tools can write passable essays, summarize cases, and generate legal arguments. Students armed with ChatGPT can fake their way through cold calls and produce work that isn't their own.
But this framing misses the larger point: AI, deployed thoughtfully, can deepen the learning process.
Consider what actually develops legal thinking. It is not the rote memorization of holdings or the ability to recite facts under pressure. It is the repeated practice of identifying issues in complex patterns, understanding how rules interact with facts, and constructing arguments that withstand scrutiny. The Socratic method works not because public interrogation is inherently valuable, but because it forces students to actively engage with material in real-time.
The question isn't whether AI belongs in legal education, but rather how we design AI tools that enhance rather than circumvent this intellectual development.
AI that generates complete answers teaches nothing. AI that provides instant, targeted feedback on student-generated work accelerates learning. AI that writes your essay undermines education. AI that identifies exactly where your reasoning went wrong and points you to the specific doctrine you misunderstood makes you better.
The legal profession itself is grappling with these questions. Major firms deploy AI for document review, contract analysis, and legal research. Tomorrow's lawyers won't be those who avoid these tools but those who understand how to leverage them while maintaining the critical thinking, judgment, and oversight that no algorithm can replicate. Law schools that pretend AI doesn't exist are preparing students for a profession that no longer exists.
Integration requires constraints: no generating original arguments, no writing essays for students, no shortcuts around comprehension. Every output should be traceable to source material. Every assistance should require engagement with underlying reasoning. The technology should create more opportunities for deep engagement with complex ideas, not fewer.
It also means being transparent about what AI can and cannot do. It can process vast amounts of information quickly. It can identify patterns and connections. It can provide instant feedback. But it cannot replace the human judgment required to navigate ambiguity, weigh competing interests, or construct novel arguments. It cannot develop the professional identity and ethical framework that law school cultivates. It cannot transform a student into a lawyer.
The path forward is not to ban AI from legal education or to embrace it uncritically. It's to establish clear principles for its use: Does this tool deepen understanding or bypass it? Does it create more opportunities for rigorous practice or fewer? Does it prepare students for modern legal practice or undermine their development? Is it accurate, reliable, safeguarded for privacy and free from hallucinations?
Law school should remain incredibly difficult. The curve should still matter. The intellectual transformation must remain at the heart of the experience. But you can't use Jobs' "bicycle for the mind" if your study tooling lets you down.
The future of legal education lies in partnership between human judgment and computational precision. Law school's ultimate test (the final exam that counts for 100% of your grade) remains entirely in your hands. No AI can sit in that examination room or construct arguments under pressure for you. What AI can do is ensure you arrive at that moment with deeper preparation and more practice than any previous generation of law students.
At Cubby, we've built our platform around this principle. Cubby fundamentally cannot be used for cheating. And we will never build products that enable shortcuts. Every response you submit is your own work. Every argument you construct comes from your understanding. Cubby ensures that understanding is as deep and precise as possible before you face the final exam that determines your future. This isn't cheating, it's advanced preparation.
This is how we lead the renaissance in legal education.