Claude for Legal: Cite Checks, Contract Deviations, Regulatory Scans
Prefer to watch? Watch on YouTube · Subscribe for new builds · Read the transcript
Anthropic just added a gallery of ready-made artifacts to Claude for Legal. These are interactive tools, like trackers and dashboards, built around the work legal teams asked for most. In this short video I go through what's in the gallery, how the tools work, and why the setup is a good model for using AI on legal work. It's for lawyers, in-house legal teams, and anyone else curious about where legal AI is going.
What I cover in this video
- 0:00: What's new: a gallery of ready-made artifacts inside Claude for Legal, built for the work legal teams asked for most
- 0:00: The cite check, which compares a legal document against its sources
- 0:00: The playbook deviation board for contracts and the regulatory horizon scan
- 0:00: How each artifact works: an example to start, then your own documents and connectors, answers linked to sources, and a lawyer making the final decision
- 0:46: Harvey and PwC have built their own artifacts for it, plus my question for you: which legal job would you give Claude first?
What the legal artifacts are
An artifact here is an interactive tool, not a one-off chat reply. Think of a tracker or dashboard you can open, use, and come back to. Anthropic put together a gallery of them for Claude for Legal, focused on the work legal teams asked for most.
A chat answer disappears into your history. A tracker or dashboard is something your team can look at together and keep using. For legal work, where you're checking the same kinds of things over and over, that's a better fit.
The three tools I walk through
Cite check
The cite check compares a legal document against the sources it cites. It flags three kinds of problems:
- Quotes that don't match the source they're attributed to
- Wrong page numbers
- Sources that don't support the claim they're cited for
That last one goes past a formatting check. It's asking whether the source actually backs up what the document says it does.
Playbook deviation board
Plenty of companies keep a set of standard contract terms, often called a playbook. The playbook deviation board checks a contract against your company's standard terms and flags every clause that differs. You don't have to read the whole agreement line by line hunting for changes. You get a board of the clauses that need attention.
Regulatory horizon scan
The regulatory horizon scan maps upcoming rules on a timeline and ranks them by impact. So instead of a flat list of regulations, you can see what's coming, when, and which items matter most to you.
How each artifact works
All the artifacts in the gallery follow the same pattern, and I think that's the most important part of this release:
- It opens with an example. Before you add anything, you can see what the tool does and what its output looks like.
- You plug in your own documents and connectors. Once you get the idea, you point it at your real material.
- Every answer links to its source. You can click through and check where a finding came from, so you're not taking the AI's word for it.
- Claude does the first pass, and a lawyer makes the call. The AI does the first round of checking and flagging. A person with legal judgment makes the decision.
- You share it across the business. Once it's set up, other people at the company can use it too.
That's why I called this the way legal AI should work. Legal work needs to be checkable, and an answer you can't trace back to a source isn't much use to a lawyer. Linking every answer to its source and keeping a lawyer in charge of the final decision gets you the speed of AI without giving up the accountability legal work needs.
Partners are building artifacts too
It isn't only Anthropic's own tools in the gallery. Harvey and PwC have built artifacts for it as well. Outside firms are making tools for this format, so the gallery isn't a one-time set of examples. Others in legal and professional services are building on it.
That leads to the question I ask at the end of the video: which legal job would you give Claude first? A cite check, a contract review against your playbook, a regulatory scan, or something else?
Key takeaways
- Claude for Legal now has a gallery of ready-made artifacts. These are interactive trackers and dashboards built for the work legal teams asked for most.
- The cite check flags quotes that don't match, wrong page numbers, and sources that don't support the claim.
- The playbook deviation board flags every contract clause that differs from your company's standard terms. The regulatory horizon scan ranks upcoming rules by impact on a timeline.
- Each artifact starts with an example, then works with your own documents and connectors, and every answer links to its source.
- Claude does the first pass and a lawyer makes the call. Harvey and PwC have also built artifacts for it.
FAQ
What are the new artifacts in Claude for Legal?
They're a gallery of ready-made interactive tools, like trackers and dashboards, built for the work legal teams asked for most. The gallery includes a cite check, a playbook deviation board, and a regulatory horizon scan.
What does the Claude cite check do?
It compares a legal document against its sources. It flags quotes that don't match, wrong page numbers, and sources that don't support the claim they're cited for.
Does Claude for Legal replace a lawyer?
No. Claude does the first pass, and a lawyer makes the final call. Every answer links to its source, so the lawyer can check it.
Who else has built artifacts for Claude for Legal?
Harvey and PwC have both built artifacts for it, alongside the ones in Anthropic's gallery.
Transcript
Read the full transcript
0:00Anthropic just added a gallery of ready-made artifacts to Claude for legal. Those are interactive tools like trackers and dashboards built for the work legal teams asked for most. So, what's on the wall? A site check compares a legal document with its sources and flags quotes that don't match, wrong page numbers, and sources that don't support the claim. A playbook deviation board checks contracts against your company's standard terms and flags every clause that's different. and a regulatory horizon scan maps upcoming rules on a timeline ranked by impact. Each one opens with an example. Then you plug in your own documents and connectors and every answer links to its source. Claude does the first pass and a lawyer makes the call. Then you share it across the business. Harvey and PWC even
0:46built artifacts for it. Comment down below which legal job would you hand to Claude first? I want to know.
Ready to start building with AI?
Join the Shipping Skool community free. Paid members get live bootcamp calls every week and weekly-updated courses.
Join Shipping SkoolKeep reading
Claude Code Just Doubled Your Limits. Here Is How I Am Using Every Last One.
Anthropic just doubled Claude Code usage limits during off-peak hours. This is how I restructured my entire AI agent pipeline to take advantage before it expires March 27th.
Anthropic is Quietly Crushing OpenAI (While Everyone Sleeps)
Anthropic hit $2.5B revenue while launching the Institute and deepening partnerships. Here's why Claude is quietly winning the AI race.
Claude Meetings Explained: Claude Joins Your Google Meet Calls
Claude Meetings is Anthropic's new beta feature, rolling out on Team and Enterprise plans, that adds Claude to your Google Meet call. It builds the doc and the whiteboard while you talk, then leaves the notes, action items, and everything it made on one page.
