Why this site exists

Most writing about artificial intelligence and the law falls into one of two camps. Vendors tell you their platform will transform your practice. Commentators tell you the profession is doomed, or that nothing will change. Neither camp shows you what to type.

LLMs for Lawyers is the third thing: a working lawyer’s field notes. Every page answers a practical question — how do I review a contract with Claude, what should I never paste into ChatGPT, which tool should my firm actually buy, how do I check whether a case the model cited exists — with concrete steps, copy-paste prompts and the caveats that only come from having been burned.

The site is the free companion to the live course AI Lab for Lawyers on Maven. The guides tell you what to do; the course makes you do it, live, in the tools, with a cohort of lawyers from around the world.

About Dr. Niklas Schmidt

Dr. Niklas Schmidt is a Partner at Wolf Theiss Attorneys-at-Law in Vienna, one of the leading law firms in Central, Eastern and South-Eastern Europe, where he heads the firm-wide tax team. He is admitted as a lawyer and as a tax adviser in Austria. Before joining the firm in 2000 he worked for several years at a “Big Four” accounting firm and as a research assistant in the tax law department of the University of Vienna. He describes himself as a lawyer, author, speaker, investor and technologist — in roughly that order on a Monday, and in the reverse order by Friday.

A few facts that matter for what you read here:

  • He has invested thousands of hours in researching practical AI applications for legal work, has read hundreds of pages of peer-reviewed research on AI, watched thousands of hours of video on it and tried out hundreds of AI apps. He keeps what he calls “an armada of AI agents” that work for him automatically every day, in business development and in his day-to-day work as a lawyer.
  • He was using AI writing tools six months before ChatGPT launched — which, as he likes to say, is his biggest flex. The story, told in his TEDxSalzburg talk “Finding a teammate in AI”, starts with a three-word text from his son — “try this out” — and a link to an obscure start-up called OpenAI.
  • He teaches AI Lab for Lawyers, a live cohort course rated 4.8/5 by 328 lawyers from hundreds of firms, including Wolf Theiss, Norton Rose Fulbright, Paul Hastings, Pillsbury, Taylor Wessing, Withers, Simmons & Simmons, Charles Russell Speechlys, Cuatrecasas and Uría Menéndez.
  • He founded the WT Crypto Academy, an internal training programme that more than 1,000 lawyers have attended and rated 4.8 out of 5 on average, and he has delivered roughly 450 presentations over 20 years, mostly at international conferences.
  • Away from AI, his practice covers corporate and private-client tax, with particular expertise in crypto assets and blockchains: he wrote a German-language reference book on the subject and co-edits the leading English-language book on crypto taxation in more than 40 countries. He sits on the advisory board of the Digital Asset Association Austria and on the editorial board of the Austrian SteuerExpress magazine, is the immediate past chair of the International Bar Association’s Private Client Tax Committee, and is a member of STEP and the International Fiscal Association. Chambers High Net Worth calls him “a phenomenal lawyer who is the best of the best in Austria”.
  • He holds a doctorate in law (Dr. iur., 2000) and a Mag. iur. from the University of Vienna, a business degree (Mag. rer. soc. oec.) from the Vienna University of Economics and Business, and completed IESE’s Advanced Management Program (2012) and a Postgraduate Diploma in Financial Strategy at Oxford’s Saïd Business School (2013). He works in German and English.

His view of the technology, from the TEDx stage: large language models “should not only be seen as tools, but rather as members of one’s team.” The guides on this site are written in that spirit — the model is a brilliant, tireless, occasionally unreliable colleague, and the skill is in managing it.

You can reach him at niklas.schmidt@wolftheiss.com, on LinkedIn, or via his Wolf Theiss profile.

How the guides are written

Each guide starts from a real task a lawyer has to do, not from a feature a vendor wants to sell. The workflow is tested in at least one general-purpose model (ChatGPT, Claude or Gemini) and, where relevant, in the legal-specific platforms. Every case, bar opinion, statistic and product claim is checked against a primary source before publication, and anything that could not be verified is left out — a site about hallucinations cannot afford its own.

The prompts are deliberately written in plain language with [square-bracket placeholders], so you can paste them and adapt them without learning any jargon. If a prompt on this site does not work for you, it is the site’s fault, not yours — tell us.

Editorial independence

This site does not run advertising and is not paid by any AI vendor. Tool recommendations are based on hands-on use and on published evidence, and they change when the evidence changes. The only thing being sold here is the course, and it is mentioned openly on every page.

The guides are educational material for legal professionals. They are not legal advice, they do not create a lawyer-client relationship, and they cannot replace your own professional judgement or the rules of your bar. Read the ethics guides before you rely on any tool in client work, and always verify what a model tells you.

Ready to practise rather than read? See the next cohort of AI Lab for Lawyers on Maven.