From Thought Leadership to Thought Doership: The Anti-AI Slop Playbook for Lawyers

August 18, 2026

The confidentiality-compatible playbook for lawyers who want a brand AI slop can’t touch.

Open LinkedIn right now and you can spot it quickly: the post that says everything and means nothing. Perfect grammar, tidy structure, three takeaways, zero fingerprints. It reads like it was written by someone who has never sat across a table from a nervous client or watched a strategy fall apart in real time. Because it was.

That flood has a name now. People are calling it AI slop, and audiences have gotten fast at identifying it. The interesting part is what’s happening in response. A growing number of brands are starting to compete on the opposite of polish. Some are publicly pledging not to use AI-generated people in their campaigns. Others are calling out the sameness of AI content directly in their ads and putting their actual process, their actual people, on display. The signal is the same one across all of them: a human made this, and here’s the proof.

This started in consumer marketing, but it doesn’t stay there. Consumer behavior sets the expectations people carry into every buying decision, including how they choose a lawyer. When a general counsel can get a competent-sounding memo out of an AI agent for free, a competent-sounding article from your firm stops being impressive. It becomes the baseline nobody is paying for.

The shift: From thought leadership to thought doership

In Harvard Business Review, John Winsor put language to what a lot of us were already feeling. His argument: generative AI has made it trivially easy to sound like an expert, which quietly kills the value of traditional thought leadership. Many have read the frameworks and sat through the keynotes, yet their organizations are still stuck. Most organizations need more than a framework or a process; they need experience that will actually guide them through the real issues affecting their business.

He calls the alternative thought doership. A thought leader tells you AI will transform your workforce, while a thought doer builds a pilot with your team in ten days, watches what breaks, and rebuilds it. One performs expertise while the other reports experience, including the parts that didn’t work. Winsor’s credibility test is blunt: if someone’s story is all wins and frameworks, they’re performing, not reporting.

For lawyers, this lands with extra force because your actual work is exactly the kind of specific, hard-won, pattern-rich experience no framework summary can replace. You are, on paper, the ideal thought doer.

There’s just one wall in the way, and it’s a real one.

The confidentiality problem (and the way through)

You can’t do what a marketer does. You can’t post the receipts, screenshot the deal, or narrate the client’s matter. Confidentiality isn’t a branding preference for you; it’s an obligation and a duty you comply with to continue practicing law.

So how do you show your work without showing your client’s work?

You separate the two things people keep collapsing. What happened to a client is confidential. How you think, work, and iterate is yours to share. Your whole approach to thought doership lives in the how, and not the who.

The playbook

1. Document your methodology, not your client’s matter

Don’t post the case. Post the reasoning you run every time you have a case like it. What’s the actual sequence you move through when you’re deciding whether to settle or litigate? What does your intake look like for a matter you’ve never handled before?

Good practitioners have a process they’ve refined over years, and they know that step one has a 1A and a 1B depending on the facts in front of them. That judgment is the credential. Show it, and you’ve proven you’ve done this enough times to know where the process bends.

Louise Mathias, litigator and mediator, does this well in her LinkedIn post from a few months ago where she argues a litigator’s job isn’t to simply assess liability and damages; it’s to ask, “Is there a narrative in this case that a jury could hijack?” You can see this type of thought leadership is demonstrating to her clients how she thinks and approaches her work.

2. Publish anonymized pattern recognition

Handle enough similar matters, and you start seeing things no single client owns. That’s an asset, and you can share it:

“Every employment dispute I’ve handled with this fact pattern in the last 18 months has settled before discovery, for the same reason. Here’s the reason.”

For example, Todd Patterson, founding partner of Patterson + Sheridan, has spent 30 years in patent law. When his firm hit its 30th anniversary, he didn’t post a milestone graphic and a thank you. He posted the pattern only three decades of watching the industry could show him.

That’s a hook, it’s true, and no one else can write it because no one else has his vantage point. Instead of recounting the firm’s history, he reports about a trend he can see only because he’s watched it play out across the field for 30 years, with the identifying details stripped out. This goes beyond a case study into intelligence. And it doubled as the most persuasive anniversary post he could have written, because it proved the firm’s whole thesis instead of just asserting it.

3. Show the strategy that backfired

This is the one AI can’t touch, and it’s the one that works.

A while back we ghostwrote a LinkedIn post for a litigator. The story was plain. Early on, he prepared the way he’d always prepared: one theory of the case, one path, the argument he was sure would win. It didn’t go the way he expected in the room, and he lost ground he didn’t need to lose. So he changed how he works. He stopped building for the single most-likely path and started preparing several at once, every time, walking in ready for the version of the day he hadn’t planned for.

We didn’t name a client. We didn’t touch a single confidential fact. It was just a real lesson, honestly told. That post outperformed every polished, on-message thing we’d ever written for him because it was the one thing on his feed that obviously came from a person who had been there.

That’s the whole move. A strategy that failed, a clause that didn’t survive due diligence, an argument you had to abandon mid-matter, sanitized of anything identifying. It’s the most valuable content most lawyers are sitting on and the least likely to post.

4. React in real time

When a significant decision drops or a regulation lands, resist the urge to wait three weeks for the polished analysis. Put the reader in the moment with you:

“I was mid-sentence drafting a clause for exactly this scenario when this opinion came down. Here’s what it changes in what I was writing, and why.”

Time-stamped, first person, unmistakably human. The opposite of slop, by construction.

Jessica Aries, our founder, uses this approach at conferences. She has posted several live reactions and immediate thoughts on conversations she’s had at conferences and has seen a great response. As we know, lawyers often are asked to attend networking events, conferences, and conventions. This is an easy way to put your client in the moment and the conversation in a brand-building way.

Here are a few examples of those posts:

Say it like a doer, not a leader

The reframe is mostly a change in language. Watch what happens when you swap the frame:

Thought leadership framing	Thought doership framing
“My perspective on the future of data privacy law”	“What I changed in my client contracts last Tuesday after this ruling”
“Key trends shaping M&A in 2026”	“The deal structure we tried in Q1 that didn’t survive due diligence, and what we replaced it with”
“Why cybersecurity is a board-level issue”	“The 72 hours after a client got a breach notice: what we did, what worked, what we’d do differently”
“Thought leadership on AI in legal services”	“Here’s the AI intake workflow I built this quarter, what broke, and what I rebuilt”
“Our firm’s approach to employment litigation”	“The one deposition question I’ve added to every harassment case since last year, and why”

The left column could have been written by anyone, including a machine. The right column could only have been written by you.

One more thing worth saying: go narrow. Broad legal overviews are the easiest thing in the world for AI to reproduce. Ask a model why a defendant settles a case and you get a tidy, generic list.

Now read how Kiernan McAlpine, a Houston plaintiff-side trial lawyer, opens a post: “Some defendants will happily lose $400,000 at trial rather than settle for $70,000. That makes no economic sense. And that’s exactly the point.”

He then explains why, from inside the machinery of insurance litigation, that irrational-looking behavior is completely rational. That’s territory a model can’t fake, and a decision-maker can’t get anywhere else, because it doesn’t come from the law. It comes from having sat across the table enough times to see what the insurer is actually optimizing for.

Make it sustainable

You have a practice to run, so build a rhythm you can actually keep:

Monthly anchor: one longer piece walking through a specific challenge you navigated, full methodology, outcome direction anonymized.
Real-time reactions, as needed: first-person commentary within hours of a decision or regulatory change, not days.
Weekly process notes: three to five sentences revealing a framework, a method, or a lesson from the actual work week.
Quarterly failure report: a deliberate look back at an assumption that proved wrong or a strategy you abandoned, and what replaced it.

The scar tissue advantage

AI has industrialized competent-sounding legal analysis. It can produce the memo, the overview, the tidy three-point post, forever, for free. What it cannot produce is the thing you earned by navigating a hard matter dozens of times and being willing to say out loud what that was actually like.

The future of legal marketing won’t belong to the firms that describe the trends best. It’ll belong to the ones documenting their craft in real time, inside the ethical lines, with the confidence that no model can fabricate scar tissue.