Field Notes from New York Estate Planners Day 2026

I recently had the pleasure of attending New York Estate Planners Day, the Estate Planning Council of New York City's annual gathering at the Union League Club. The lineup was a thoughtful one, ranging from successor trustee complexities and §672(c), to how to actually implement AI in your practice, to protecting a will from attack, to the fiduciary accounting and tax changes coming down the pike.

I was able to catch two of the sessions in person: Marty Shenkman on practical AI, and Stephen Liss on successor trustees. A few threads from those conversations have stuck with me since.

The Rising Floor of Practitioner Competence

The floor on practitioner competence is rising fast, and Marty framed it most directly. Staying current with AI is itself becoming an ethical obligation, and one that lawyers cannot outsource to their tech department or colleagues. The technology competence rules have been on the books in most states for years, but they land differently now that the tools are just a browser tab away.

Confidentiality and privilege come first. No client information in public APIs, ever. But beyond that baseline, the practitioners who treat AI as part of their craft are going to pull away from the ones who treat it solely as a back-office tech matter.

The Highest-Leverage Use of New Tools Is Attention

The highest-leverage use of new tools, in Marty's framing, is being more present with your client.

His approach is simple. He records his client meetings with consent, then lets AI handle the transcript and the follow-up memo, which lays out what was discussed and the next steps. The mechanical work that used to compete for his attention during the meeting now happens after the fact, on its own.

What he described, though, isn't really about faster note-taking. It's about no longer multitasking through the conversation itself - no longer scribbling action items while a client describes a family matter that may shape the whole plan. He just listens.

That reframing stuck with me. What these tools really give back is attention - and in this work, attention is almost always the thing in shortest supply.

The Harder Problem: AI on the Plan, Not Just the Document

The meeting is one piece of it. The bigger lift, and the one that's harder to crack as a practitioner, is how to leverage AI with the plan itself.

You can upload a trust to a generic LLM today and get back a competent summary. That's useful, as far as it goes.

The harder problem is what comes after. Serving a family well means keeping a working view of the plan as amendments get layered on, trustees change, families grow, and the tax landscape shifts underneath. A one-time read of a one-time summary doesn't carry you through that.

That's the part of the problem that's still wide open in most practices. The documents sit in one folder, the summary in another. Six months later, when something changes, the attorney is back to re-reading the underlying instruments to remember what the plan actually says.

That's the part Luminary is built for: turning the documents into structured and dynamic data the advisor and attorney can actually keep working with, rather than re-summarizing the same instruments every time a question comes up.

Successor Trustees and the Mechanics That Quietly Matter

Stephen Liss's session was a useful reminder that successor trustee design deserves more attention than it usually gets. His central argument was that the common practice of barring all §672(c) parties from serving as successor trustees over estate tax-exempt irrevocable trusts - a convention that grew out of Rev. Rul. 95-58's safe harbor - is more restrictive than the underlying tax rules actually require, and that practitioners may have more drafting room than the convention suggests.

That drafting room matters because the choice of trustee shapes more than people often realize. Grantor trust status, dispositive provisions, decanting authority: all of these turn on who is actually sitting in the trustee seat when the original named trustee can no longer serve.

The result is that the underlying mechanics can quietly determine whether the plan still works as intended years down the road. A successor steps in, the drafter is no longer around to ask, and whatever assumptions were baked into the original design either hold up or they don't.

Stephen's session reinforced for me that this is one of the places where careful drafting at the front end saves a meaningful amount of remediation later. It is also one of the places where a structured view of the plan (who can do what, under which conditions, and with which consequences) earns its keep over the long arc of the family.

Where the Practice Is Heading

Taken together, these threads map the shape of where the practice is heading. Some of it is technology, and some of it is the careful drafting and design work that has always mattered.

The practitioners who come out ahead over the next few years are going to be the ones who let the right tools take the mechanical work off their plate, freeing them to spend more time on the substantive questions that define a good plan. I came back from Estate Planners Day with a clearer sense of what those questions look like, and a renewed appreciation for the colleagues thinking carefully about them.

These are also the problems we work on every day at Luminary: how to give an attorney time back for the work that actually requires their judgment, and how to make the mechanics of a plan visible at the moments when they actually shape outcomes. Structured data and a living view of the plan give an advisor's judgment something to build on, rather than starting fresh every time the file gets reopened.

If you're interested in learning more, you can request time with a Luminary expert here.