By: Christopher D. Warren

Moral Machine Setlist: Bob Dylan, “The Times They Are a-Changin’” (Listen on YouTube)

The steamer in the picture above has a name on her hull, Legacy Law Firm, and she is going down. The smoke is real, the list is getting worse, and the people who can read the water are already finding their way off. Some of them are reading it right now.

Staying aboard a firm that refuses to modernize is not loyalty. It is a risk, and for the lawyers running that practice it is starting to look like something that has a name in tort law. The name is not “tradition.” It is liability, measured against a standard of care that no longer waits for the holdouts. And there is usually one hand at the wheel who will not turn it, the senior partner who mistakes the consensus of their own vintage for principle and makes everyone below them sign for it. We will get to them.

Competence is not a hobby anymore

The duty of competence is old. What counts as competent is not. The ABA amended Comment 8 to Model Rule 1.1 back in 2012 to require that a lawyer keep abreast of “the benefits and risks associated with relevant technology.” Most states have since adopted some version of it. New Jersey issued its own preliminary guidance on attorney use of AI in January 2024, and the ABA’s Formal Opinion 512 put competence at the front of the whole analysis in July 2024.

Notice the word the holdouts skip. The duty runs to benefits, not only to risks. You cannot weigh a benefit you have refused to learn. You would not see a doctor who works only from memory when the literature is one search away, and your clients are owed no less. The first lawyers trained on these tools from their first day of law school are already on the way, and the competence gap is no longer a prediction. It is closer to a graduation date.

And here is what the holdouts will never let themselves find out: the learning is fun. I spent a weekend building my own harness, wiring a small language model into a chain of API calls and letting the rig run a genuinely agentic, multi-step task while I made coffee. And honestly, the friction is the fun part. Marcus Aurelius beat me to it eighteen centuries ago, in his Meditations: “The impediment to action advances action. What stands in the way becomes the way.” I’m pretty sure he was talking about AI.

Necessary for my practice? Of course not. Better informed about what these systems can actually do? Without question, and that is the whole point. Whether you are advocating for a client or guiding one as they build AI into their own business, or preaching to the room about its evils, doing any of it without ever diving into how the thing works is astronomy from inside a cave. You can use all the right words, recite every constellation, and still have never once seen the sky.

The bill is the other half

Competence is only half of why the holdout is exposed. The other half is the invoice. RPC 1.5 requires a reasonable fee, and its factors are tied to real inputs, among them the time and labor required and the fee customarily charged for similar work.

The billable hour pays for time, so once AI turns an eight-hour memo into a one-hour memo, the only way to bill the old figure is to charge $5,000 or $6,000 for the single hour that remains. No client on earth signs that engagement letter. The value of the work did not shrink. It grew. Only the hours fell, and the hour can no longer carry the price without a rate that is absurd on its face. That leaves two roads. One abandons the billable hour for an alternative fee arrangement. Opinion 512 expressly blesses the flat-fee and subscription structures that solve this. The other road keeps the hour and bills for time that was never spent. That is not a pricing model. It is fraud, the very thing RPC 8.4(c) makes professional misconduct. That’s the opposite of what an attorney is obligated to do.

Yes, it hallucinates. That is the point.

The standard objection writes itself. The lawyers getting sanctioned are not the ones avoiding AI. They are the ones who filed its output without reading it. That is fair. There are now more than 1,600 documented incidents of fabricated AI citations in court filings, the overwhelming majority of them from 2025, and a federal court in Oregon recently set the record with a sanction near $110,000.

Those cases do not make avoidance the safe harbor. They prove the opposite. The duty is competence, and competence includes understanding how the tool fails and checking its work before it leaves your desk. Competent practice pairs the reach of the technology with the judgment of a lawyer who reviews everything it produces.

The part that should actually worry them

No court has held that simply declining to use AI is, by itself, a basis for liability. Liability of that kind is a tort, and a tort requires a duty, a breach, causation, and actual harm, and a disciplinary violation of the RPCs is a different question from civil liability. A lawyer who never touches these tools and never harms a client is not getting sued for that alone.

The standard of care is where the real exposure lives. It is not fixed by statute. It is set by what reasonably competent practitioners actually do, and that benchmark is moving quickly. A lawyer who has fallen behind on the tools while overcharging for the slow version of the work is drifting under a standard that keeps rising.

And the standard is not only drifting in the background anymore. Some of it is now in writing. On June 1, 2026, Part 161 of the Rules of the Chief Administrator took effect across every court in the New York Unified Court System, in civil and criminal cases alike. It rewards a close reading, because it is more interesting than the headlines made it sound.

Part 161 does two things system-wide, and neither one is a ban. It says the use of AI tools in preparing court papers should not be prohibited, and it says lawyers are not required to disclose that they used them. Then it supplies a Model Rule, in Appendix A, that individual courts may adopt as their own.

Use it, you need not even announce it, but you own every word you put your name to. And even in a court that has not formally adopted Appendix A, the floor was already there: by signing a paper you certify, under 22 NYCRR 130-1.1 and the duty of candor in RPC 3.3, that it holds no false statement and no frivolous argument. The signature is the checkpoint, and it cannot be handed off to a junior associate, a vendor, or the model.

The captain who won’t leave the bridge

Every sinking firm has one. They are the senior partner with thirty years of wins behind them, who built the firm’s name on craft and late nights in the library. They believe the slow, hand-built way is not just their way but the honorable one. Their refusal is fear and ego in the robes of principle, and that is what makes it dangerous. It is not theirs alone: it is ratified every week by a room of partners their own age, all nodding at the same refusal until a generational reflex hardens into what they mistake for consensus.

You know their lines, because they sound like wisdom in a partner meeting and read like an epitaph years later. I practiced law for thirty years without a computer telling me how. We are a relationship business, not a technology company. When I was an associate…. And so on. Each line buys them another quarter of standing still and collecting a check while the standard of care keeps moving without them.

Here is the part that turns them from a relic into a unintentional villain. A legacy leader who refuses these tools and quietly retires harms no one. This one does worse. They force the refusal downward. They ban the tools across the firm and bury approvals in bureaucracy for years, then demand associates bill 2200 hours a year. They keep their name off the filing and make sure the Junior Partner’s name is on it. They collect the profit of the hourly model and park its risk on a younger lawyer’s license. They ordered the memo. They signed nothing. The certification, and the exposure, travel down to the person they told to do the work the way they did it 35 years ago.

They tell the room they are the steward, managing risk, the steady hand on the wheel. They are in fact the single largest source of risk in the building. They drive the talent out the door, drive the tools into the shadow AI the firm pretends it has banned, and hand every competitor a structural head start in every matter that follows. They think they are the lighthouse, right up until the hull opened.

When the firm is the problem

Here is where it stops being abstract, and where the ship starts taking on water faster than the people aboard understand.

A firm can ban AI. Plenty have, on the theory that a prohibition is the safe move. It is not the safe move, and Part 161 helps explain why. A firm that bans them outright has planted itself somewhere the courts specifically refused to stand, and it has done so at its own lawyers’ expense.

Because a ban does not stop anyone from using AI. It only drives the use underground. The associate buried at midnight still opens a consumer chatbot on a personal phone, with no governance, no privilege protection, and none of the firm-sanctioned tooling that was actually built for legal work. That is shadow AI, and a prohibition is the most reliable way to manufacture it.

So if your own firm is the thing standing between you and a modern competent practice, take it seriously, because the standard of care, and now the signature on your filings, will be measured against what good lawyers do and not against what your firm’s policy happens to permit. Firms that took this seriously early already exist, and in the interest of disclosure, mine is one of them.

The water is already at the deck. The lawyers who can read it are already in the lifeboats. You can call that disloyalty, or you can call it competence, because they are now the same thing. The tide will not argue the point. Read the water, or go down with the ship.

If you are looking for a firm that has already embraced the modern practice of law, and you are ready to leave the sinking legacy models behind, come find us. We are always looking for lawyers who can read the water. A lifeboat keeps you afloat; an airship leaves the water behind. There is room on ours, and it is climbing.

DISCLAIMER: This summary is not legal advice, and does not create any attorney-client relationship. This summary does not provide a definitive legal opinion for any factual situation. Before the firm can provide legal advice or opinions to any person or entity, the specific facts at issue must be reviewed by the firm. Before an attorney-client relationship is formed, the firm must have a signed engagement letter with a client setting forth the Firm’s scope and terms of representation. The information contained herein is based upon the law at the time of publication.

Views expressed on Moral Machine are the author’s own and do not reflect those of the New Jersey Supreme Court Attorney Ethics Committee (District VI) or Falcon Rappaport & Berkman LLP.