I built a legal research agent that writes structured memos in 3 minutes.

I built a contract review tool that flags 23 risk categories across any commercial agreement.

I built a client intake system that cut my admin work from 3 hours a week to 20 minutes.

None of this required a developer. None of it required coding knowledge. I built all of it using AI tools that are available to every lawyer and law student right now.

And most of you have no idea these tools exist.

That is the problem I am solving on April 18th.

I am running a free 2-hour live workshop where I will show you exactly how AI can transform the way you practice law, study law, and build a legal career.

This is not a webinar where someone reads slides about "the future of legal tech." I am going to open my laptop, share my screen, and build things in front of you. Live.

Here is what I will cover:

- How to build AI agents that handle legal research, drafting, and document review in minutes instead of hours. Not in theory. I will build one live during the workshop.

- How to automate the repetitive parts of legal practice: client intake, memo generation, contract checklists, compliance tracking — so you can focus on the work that actually requires a lawyer's judgment.

- The exact AI tools I use in my own practice every day. What works. What does not. What is worth paying for and what is a waste of money.

- How law students can use AI right now to prepare better research, write better memos, and walk into their first job with skills most senior associates do not have yet.

- Prompt engineering for legal work, the difference between getting generic output and getting something you can actually send to a client.


The details:

Date: April 18, 2026
Time: 8:00 PM IST

I am keeping registrations limited because I want this to be interactive, not a broadcast. If you have questions during the session, I will answer them live.

Register here: https://lnkd.in/gnzEcQ4s

Who should attend:

- Law students who want to graduate with skills their peers will not have.
- Junior associates drowning in research and drafting work.
- Solo practitioners and freelance lawyers who want to take on more clients without burning out.
- Senior lawyers who keep hearing about AI but have not seen it actually work on legal tasks.
- Anyone who has tried ChatGPT for legal work and thought "this is not useful", because you were using the wrong tool the wrong way.

I have spent hundreds of hours testing AI tools on real legal work. Contracts, memos, research, compliance, client communication. I know what works and I know what fails. This workshop is everything I have learned, compressed into 2 hours, demonstrated live.

See you there.

#AIforLawyers #LegalTech #LawStudents #LawyersofLinkedIn #LegalAI #FreeLegalWorkshop #AIWorkshop #LegalInnovation #FutureLawyer #AIAgents

You see your colleagues using AI in their in-house legal work and want to do the same.

Exactly how to do that, though? That part can feel insurmountable, especially if you put high pressure or high stakes on the result of what you hope to use it for.

Join Kevin Keller and me, along with the Sandstone team, next week to talk about how we're using AI in our legal departments.

I'll focus on what has and hasn't worked in my own journey to using AI for my in-house legal work. My use cases are simple, but the impact is significant. I'll take these steps farther one day, but the value I've seen to date is worth all the time and energy to get to where we are.

On my end, I'll discuss:
- Where dedicated legal tech didn't work for us
- Building an NDA review agent to act like a fellow teammate (but still encourage human review)
- Trying, failing, and learning on what successful legal matter tracking looks like for us

See you April 9th! Register in the link in comments.

𝗔𝗜 𝗦𝗼𝘂𝗻𝗱𝘀 𝗟𝗶𝗸𝗲 𝗮 𝗟𝗮𝘄𝘆𝗲𝗿. 𝗧𝗵𝗮𝘁'𝘀 𝘁𝗵𝗲 𝗣𝗿𝗼𝗯𝗹𝗲𝗺.

I was a Prodigy kid.

Not the musical kind. The dial-up, message-board, pre-AOL kind.

My fascination with tech never waned. It took me to the White House, where I helped shape early federal cloud policy, launch the U.S. Digital Service, and advance the government's first bug bounty program. I saw what it looks like when transformative technologies move from experimentation to adoption at scale.

So when generative AI arrived, I wasn't skeptical. I was excited. I still am. But I'm also a lawyer. And that's where it gets complicated.

𝗧𝗵𝗲 𝗣𝗿𝗼𝗺𝗶𝘀𝗲 𝗜𝘀 𝗥𝗲𝗮𝗹

AI has real utility in legal work. My team uses it for the first redline of routine documents. NDAs, vendor agreements, standard commercial terms. It saves time and catches obvious stuff. As a starting point, it's useful. Key emphasis on "starting point."

𝗧𝗵𝗲 𝗣𝗿𝗼𝗯𝗹𝗲𝗺 𝗜𝘀𝗻'𝘁 𝗧𝗵𝗮𝘁 𝗔𝗜 𝗜𝘀 𝗪𝗿𝗼𝗻𝗴. 𝗜𝘁'𝘀 𝗧𝗵𝗮𝘁 𝗜𝘁 𝗦𝗼𝘂𝗻𝗱𝘀 𝗦𝗼 𝗥𝗶𝗴𝗵𝘁.

AI has mastered the aesthetic of legal analysis without mastering the substance. The output looks like a memo from a senior associate. Confident, structured, occasionally citing cases that don't exist or don't support what the model claims.

I've seen AI-generated redlines come back and I can tell immediately. Provisions that look like standard market terms but don't carry settled legal meaning. Indemnities that shift liability in ways the drafter didn't intend.

"𝗕𝘂𝘁 𝗖𝗵𝗮𝘁𝗚𝗣𝗧 𝗚𝗮𝘃𝗲 𝗠𝗲 𝗮 𝗗𝗶𝗳𝗳𝗲𝗿𝗲𝗻𝘁 𝗔𝗻𝘀𝘄𝗲𝗿"

I hear this more and more. I don't take it personally. Efficiency is a legitimate goal. But there is a real difference between information and legal judgment.

AI cannot know your company's litigation history, your board's sensitivities, or the regulatory posture you've taken in past filings. Law is not a lookup table. It is applied judgment under uncertainty, shaped by facts specific to you.

𝗧𝗵𝗲 𝗤𝘂𝗲𝘀𝘁𝗶𝗼𝗻 𝗖𝗼𝘂𝗿𝘁𝘀 𝗛𝗮𝘃𝗲𝗻'𝘁 𝗔𝗻𝘀𝘄𝗲𝗿𝗲𝗱 𝗬𝗲𝘁

Courts know how to interpret contracts, but they have not yet confronted AI-drafted clauses at scale.

What happens when a court is asked to interpret language optimized for fluency rather than precision? Will litigants discover the clause they're fighting over does not carry the meaning they assumed?

That moment is coming. Tech moves fast; contract disputes do not. AI-generated clauses will spread long before courts are asked what they actually mean.

𝗪𝗵𝗲𝗿𝗲 𝗗𝗼𝗲𝘀 𝗧𝗵𝗮𝘁 𝗟𝗲𝗮𝘃𝗲 𝗨𝘀?

Use AI as a starting point. Train it on your playbooks. Let it handle the first pass. But have a real lawyer review it. Someone who knows your company, your risk appetite, and the enforceability questions that don't show up in a training dataset.

AI produces information. Lawyers provide judgment. Right now, you need both.

𝗜'𝗱 𝗹𝗼𝘃𝗲 𝘁𝗼 𝗵𝗲𝗮𝗿 𝘆𝗼𝘂𝗿 𝗽𝗲𝗿𝘀𝗽𝗲𝗰𝘁𝗶𝘃𝗲.

A year ago, I shared on LinkedIn that contract drafting was becoming obsolete. The post sparked plenty of disagreement, including from Ken Adams (I always appreciate good debates).

My next prediction: contract negotiation, as lawyers know it today, is about to change too.

Contract negotiation never took too long because lawyers typed too slowly.
It took time because legal judgment depends on context.

When I was in-house counsel, it was completely normal to revert on contract redlines in a week. Sometimes longer. This was because, before you could take a sensible position, you often needed to understand what the business actually cared about, what had been agreed before, where the fallback positions were, and which precedent was actually relevant.

The first wave of legal AI helped lawyers generate and edit text faster.

What is changing now is that AI systems are starting to surface the context that sits behind legal judgment and negotiation strategy.

➡️ In this new episode of the Legal Intelligence Platform (LIP) deep dive, I explore how I would use an AI contract assistant, SimpleAI by SimpleDocs, to review redlines and negotiate a contract as in-house counsel today, and share my pov on where I think the real upside and limits still are.

✴️ Watch here: https://lnkd.in/gQ8P96Zt
---
📌 Stay tuned for part two (dropping tomorrow), where I ask Preston Clark, CEO of SimpleDocs, some tough questions about:
- Why he thinks SimpleDocs can win in contract AI, the most crowded category in legal AI, and
- How legal teams should evaluate products in a market full of "lookalikes".

Note: This video was made in partnership with SimpleDocs. SimpleDocs provided product access and reviewed the video for factual accuracy. Editorial opinions and conclusions remain mine.

Last year I started tracking how I spent the first two hours of work. Almost all of it was sorting. Emails, requests, figuring out what was actually urgent. Real legal work didn't start until 10am most days.

So I built something to change that. A few weekends of Claude prompts and some basic automations. The whole setup costs less than a steak dinner.

By the time I open my laptop now, everything that came in overnight has already been tagged. Here's what the color coding means:

Green: AI handles it completely. A standard NDA comes in and matches our playbook. If no unusual terms, AI drafts the response, marks it ready to send. I spot-check a few per week.

Yellow: AI gets it 80% done. A counterparty sends an agreement with non-standard limitation of liability terms. AI redlines against our playbook, writes the initial response, flags the 2-3 issues that need my actual judgment. Usually 10 minutes of work instead of an hour.

Red: Needs my brain. IP renegotiation on an existing deal. A transaction structure question from the CFO. Something where there's no playbook answer. AI can pull context and prior examples, but the call is mine.

Gray: Not actionable today. Industry updates, informational stuff. Filed and out of the way.

What made this actually work was realizing that most of the weight of being a GC isn't the legal judgment. It's the operational layer around it. Tracking what's due, figuring out what's urgent, staging docs for calls. That used to eat my mornings. Now it doesn't.

I didn't build this because I saw someone else do it. I built it because I needed to leverage my mornings when my brain is fresh to tackle the complex problems instead of doing ops.

If you run a legal function and you're still manually triaging your inbox at 8am, it's worth asking whether you actually need to be.

What's the one task you do every morning that you know could be automated?

#inhouselegal #legaltech #AIforLawyers

I spent 30 days building a massive AI contract management system. Then I built a standalone agent in 2 days that did 80% of the same work.

When Claude’s Legal plugin was released, the industry was shocked. But after testing it myself, I realized the "fastest path" to a working AI agent has almost nothing to do with code. It’s about capturing the judgment calls an expert carries in their head.

Most guides start with the technology. Pick a model. Choose a framework. Define a stack. Wire up an API. That's starting a house with the roof.

But a senior legal advisor doesn't review contracts by reading every word. They check 24 specific things against an acceptable range. That's the decision framework your agent needs.

𝐌𝐲 𝐛𝐢𝐠𝐠𝐞𝐬𝐭 𝐥𝐞𝐚𝐫𝐧𝐢𝐧𝐠𝐬 𝐟𝐫𝐨𝐦 𝐛𝐮𝐢𝐥𝐝𝐢𝐧𝐠 𝐚𝐧 𝐀𝐈 𝐋𝐞𝐠𝐚𝐥 𝐀𝐠𝐞𝐧𝐭:
1. 𝐃𝐞𝐟𝐢𝐧𝐞 𝐧𝐨𝐭 𝐣𝐮𝐬𝐭 𝐖𝐡𝐚𝐭, 𝐛𝐮𝐭 𝐖𝐡𝐨. Frame the role: "You act as a contract analyst, not legal counsel. You represent the seller. You surface findings—humans decide." This single sentence prevented more errors than pages of instructions.
2. 𝐆𝐢𝐯𝐞 𝐭𝐡𝐞 𝐚𝐠𝐞𝐧𝐭 𝐠𝐫𝐨𝐮𝐧𝐝 𝐭𝐫𝐮𝐭𝐡𝐬. AI needs absolute guardrails: "Unlimited liability is never acceptable." Establish these early so the agent doesn't "reason" its way into a plausible-but-wrong conclusion.
3. 𝐀 𝐩𝐫𝐨𝐦𝐩𝐭 𝐬𝐩𝐞𝐜 𝐢𝐬 𝐧𝐨𝐭 𝐚𝐧 𝐞𝐧𝐠𝐢𝐧𝐞𝐞𝐫𝐢𝐧𝐠 𝐬𝐩𝐞𝐜. The vocabulary optimized for a developer is not what an AI model needs. Write your prompt spec to be self-explanatory in full.
4. 𝐍𝐞𝐯𝐞𝐫 𝐚𝐬𝐬𝐮𝐦𝐞 𝐭𝐡𝐞 𝐚𝐠𝐞𝐧𝐭 𝐰𝐢𝐥𝐥 𝐢𝐧𝐟𝐞𝐫. If you think something is "obvious," write it down. Explicitness is the baseline requirement for reliable behavior.
5. 𝐑𝐞𝐩𝐞𝐚𝐭 𝐜𝐫𝐢𝐭𝐢𝐜𝐚𝐥 𝐜𝐨𝐧𝐭𝐞𝐱𝐭. Don't assume Module 4 remembers the "Who" from Module 1. Repeat the most critical guardrails at every stage.

The bottleneck was never the AI. It's getting a domain expert to decompose "this feels risky" into specific, testable conditions. That's the hard work. Everything else is configuration.

What domain are you thinking about building an agent for?

As a “tech-savvy” lawyer, this is how I use AI — and where I refuse to rely on it.

It’s important for us, as lawyers, to understand that AI is not intelligence, it is prediction.

It does not understand law. It recognizes patterns in how law has been written, argued, and interpreted before. This distinction matters more than most people realize.

In my practice, AI has become an instrument of acceleration — but never a substitute for judgment.

- I use AI to interrogate large volumes of information quickly.
- To identify structural patterns across agreements.
- To compare regulatory approaches across jurisdictions.
- To test the internal consistency of legal reasoning.

It compresses hours of mechanical effort into minutes. But as lawyers, it’s important for us to understand that law is not a mechanical profession, it is a profession of consequence.

AI can tell you what is typical. It cannot tell you what is safe.

It can identify what has been done before. It cannot evaluate what should never be done.

It does not bear liability.
It does not exercise fiduciary responsibility.
It does not understand risk in the way a human advisor must.

The greatest risk of AI in law is not that it will be wrong. It is that it will sound right. Create a nice attractive graphic also for this list

Which is why the real shift is not technological. It is cognitive.