Technology’s Effect on Advocacy
Five Hours for a Sheet of Pleading Paper: What I Learned When I Stopped Waiting for the Legal Tech Industry to Save Me By Benjamin Schenk My pleading paper problem
I closed the laptop, walked out to the patio, and made a decision: I was going to learn how to build software. Not because I wanted to become a software engineer — I am a trial lawyer, and I love being a trial lawyer — but because the gap between what legal tech was selling and what my firm needed had become embarrassing. I was embarrassed. That was six weeks ago (as of writing this).
Last spring, my paralegal and I spent five hours trying to format a single sheet of California pleading paper. We had just tried an AI product off the shelf. One vendor sold us by saying "we will generate discovery requests for you" — but could not get them onto pleading paper. Another's entire pitch was "we get them on pleading paper." That was the state of the art. Not a brief. Not a motion. The template . The line- numbered left margin required by Rule 2.108, the firm caption, the attorney block, the footer that locks to the bottom of the page. That crisp, presentable document reflecting the thought and effort that went into creating it. This is the format every California Superior Court litigator has fought with for decades. Every plaintiff firm I know has its own slightly-broken version, passed around like a folk remedy — you learn which paralegal has the cleanest copy and ask for it like an old family recipe. Five hours. For a sheet of paper.
Why we built instead of bought
I will tread lightly here, because legal tech has built some genuinely impressive AI products. But we built rather than bought for a simple reason: I wanted The Schenk Law Firm to own its code, every template, every workflow, every line, not rent a black box. Almost every plaintiff-side AI product shares the same architecture: ingest the firm's prior work product (complaints, demands, discovery, briefs), index it, and surface the right template at the right moment. When it works, the model sounds like the firm because it has read the firm's work. But the pipeline depends on a precondition the marketing glosses over: the prior work must be organized so AI can actually use it. Three thousand briefs in a shared drive give an AI almost nothing, a pile of unlabeled files. To retrieve "the meet-and- confer letter from that deposition-scope dispute in San Diego County in 2023," someone first has to tag every document along the dimensions that matter including case type, dispute type, county, and outcome. This tagging is an index, the card catalog that finds the right document without opening every
Benjamin Schenk is a co- founder of The Schenk Law Firm, LLP, a plaintiff firm based in San Diego, California. He is a graduate of Dartmouth College and the University of San Diego School of Law.
ben@schenklawfirm.com
66
Consumer Attorneys of California
FORUM September/October 2026
Made with FlippingBook - Online catalogs