AI finds answers. Sometimes right, sometimes wrong. Good lawyers find the right questions. Always.
This site covers two topics:
- Best Practices with AI Tools – Using AI effectively for legal work.
- AI in Contracts – managing AI-related issues in contracts
With more than 25 years of experience as a lawyer in multinational companies (IBM and Syngenta) focusing on multinational contracts, litigation and environmental matters I have developed a particular interest for AI. The “AI for lawyers” playlist of my Youtube channel covers best practices for lawyers in using AI tools as well as considerations around how to cover the use of AI in contracts.
DISCLAIMER: The views expressed on this website are my personal views and reflect the knowledge and tools as per the date of the recording.
recorded June 17, 2025
Materials related to the Video “Basic AI training for lawyers” which covers generic prompting tipps and the basic uses cases, i.e. drafting, review, summarizing, analysis / matter strategy, translation, information searching / research):
Training video: https://youtu.be/xKA-fOe5-DM?si=pTLP5aNNrP71TlsJ
Cheat Sheet (with sample prompts): 20260223 Basic AI Use Cases for Lawyers – Cheat Sheet
ACC toolkit: https://www.acc.com/resource-library/artificial-intelligence-toolkit-house-lawyers
Recorded March 1st 2026
A Governance Model for Leveraging the Benefits While Mitigating the Risks of AI Use by Lawyers.
“A danger known is a danger avoided.”
AI tools have transformed legal work — improving quality, speed, and efficiency across drafting, reviewing, and analysing documents. But even the most sophisticated models hallucinate. They produce fundamental errors that follow no pattern, appear without warning, and cannot be reliably caught by spot checks.
This video is for lawyers who already use AI regularly and want to move beyond the basics. It covers four things: the structural reasons why AI errors occur and why they will persist regardless of tool sophistication; why these errors are particularly difficult for lawyers to detect; a practical risk-based governance model for defensible AI use; and a structured approach to client transparency — including when disclosure of AI involvement is required and when it is not.
The framework presented is consistent with ABA Formal Opinion 512 on Generative Artificial Intelligence Tools (July 29, 2024).
Video: https://youtu.be/Kk7ENH7iwKw?si=nxP-VmY6v0mTJK5Z
Slide deck: 20260301 AI_Risks for Lawyers
Checklist: 20260301 checklist AI errors for lawyers
ABA opinion 512: aba-formal-opinion-512 (or here directly from the ABA website)
Recorded March 29 2026
AI tools can help lawyers to handle complex, structured legal tasks with remarkable speed and accuracy. But only if you teach them to think like a lawyer.
This video shows how to build a reusable AI workflow for recurring, structured, reference-based legal tasks — without any coding.
The example used throughout is a supplier contract review against your own template, group standards and policies.
It covers five steps:
Prepare — upload the reference documents and instruct AI to internalise them
Check — verify document completeness of the documents to be reviewed, and determine hierarchy before analysis begins
Review — compare supplier position against the reference
Output — populate a pre-defined structured assessment template
Validate — human review
The workflow in this video builds directly on the prompting principles from the Basic AI Training video and applies the governance model from the AI Errors video — in particular the mandatory human validation step that closes every workflow.
Video: https://youtu.be/REYAJYPtmuQ?si=iX1HUr6GHWW2D8oZ
Slide deck: 20260328final AI_Workflows_for_Lawyers
Assessment Template: 20260328_Generic AI_assessment_template
Workflow Instructions: 20260327 workflow_instructions_v2
Recorded April 16, 2026
Most AI markup tools work as a black box: input goes in, output comes out, you fix what’s wrong at the end.
This one is different. The model has four steps and two players: you and AI: 1. AI spots the gaps. 2. You decide how to handle them. 3. AI creates the markup. 4. You validate.
The key difference to other approach is, that instead of creating a markup for each deviation, AI proposes a concept — not specific wording — for how to address it. Then you decide: proceed as proposed, skip it, or go a different direction. That initial steer, given before a single tracked change is written, substantially improves the quality of the output compared to standard markup tools.
This video shows how to build this supervised markup workflow for complex, recurring supplier contracts reviewed against your own templates, group standards and policies and general legal and contractual knowledge .
Video: https://youtu.be/qHMj5fVynk4
Slide deck: 20260417 External supervised_ai_markup_video final
Instructions: 20260417_instructions_markup_video_updated
- Dealing with supplier use of AI -> impact on representations and warranties (incl minimum standards and checklist)
- Right tool for the right task
- advanced promtping techniques