
"In a recent episode of "Notes to My (Legal) Self," Linsey Krolik, a law professor at Santa Clara University and longtime in-house counsel at companies like PayPal and Arm, made a compelling case for what she calls AI literacy. But the bigger insight was between the lines: you can't use AI effectively in a legal setting without understanding the inputs, and that means contracts."
""We are using generative AI today, whether we want to admit it or not," Linsey said during the interview. "It's happening. So get on board and we can learn together." That sense of collective learning, and the gap between curiosity and confidence, is something many in-house teams are experiencing firsthand. There is pressure to move quickly, reduce turnaround time, and do more with less. AI promises all of that. But as Linsey pointed out, we need to start with the basics."
"She's training law students to build real-world legal documents like terms of service and privacy policies for early-stage startups. These students are already experimenting with AI. They're learning where it helps, where it fails, and how to critically assess its output. They are developing muscle memory not just in drafting, but in understanding why contracts are structured the way they are. That foundational skill, contract literacy, is what too many practicing teams are missing."
Generative AI is already being used in legal practice and teams are under pressure to accelerate work and cut turnaround time. Many teams jump straight to tool selection and output trust without auditing the inputs. Contracts and contract data determine AI behavior, and messy contract data produces unreliable AI results. Training that combines drafting real-world contracts with AI experimentation builds critical skills in assessing where AI helps, where it fails, and why contract language exists. Developing contract literacy enables safer AI adoption by improving inputs, reducing hallucinations, and aligning tool use with legal judgment and risk management.
Read at Above the Law
Unable to calculate read time
Collection
[
|
...
]