How AI Will Reshape the Practice, the Role of the Partner, and the Future of the Associate
Artificial intelligence is moving rapidly from an interesting technology experiment to a fundamental component of professional practice. For lawyers, the question is no longer whether artificial intelligence will affect the practice of law — it is how quickly, how deeply, and in what areas.
Tax and estate planning are particularly interesting in this respect. At first glance, they appear highly vulnerable to artificial intelligence: much of the work involves statutes, regulations, cases, administrative guidance, tax returns, financial information, document drafting, research, and analysis, all areas in which AI is becoming increasingly capable. But that does not mean tax and estate-planning attorneys will become obsolete. Quite the opposite may be true.
The practice is likely to divide into two very different categories: routine legal production will increasingly become automated or dramatically accelerated, while sophisticated planning, judgment, client counseling, business strategy, and relationship management will become more valuable. The result may not be the disappearance of the tax and estate-planning lawyer, it may be the transformation of the lawyer into something closer to a strategic architect of wealth, business, and family planning.
I. AI Is Already Reshaping the Practice
The transition is already underway. Legal-specific AI platforms are moving well beyond simple chatbots. Thomson Reuters’ CoCounsel, for example, is designed to combine legal research, document analysis, drafting, and increasingly agentic workflows with Thomson Reuters’ legal content and the Microsoft 365 environment. Thomson Reuters reported in 2026 that 41% of law firms surveyed were using generative AI, compared with 28% the prior year.
Harvey has developed another important model: a platform designed specifically for professional legal work, including research, analysis, drafting, and complex workflows. Its rapid growth illustrates the amount of capital and institutional confidence being placed behind AI built specifically for lawyers. In September 2026, Harvey announced a $550 million financing round at a reported $15.5 billion valuation. Other platforms, including LexisNexis’ AI products and emerging law-firm-specific systems, are similarly attempting to integrate authoritative legal information with generative AI.
The significance is not simply that a lawyer can ask AI to “write a memo.” The more important development is that AI is increasingly able to work across large collections of information and perform multiple steps in sequence. That distinction matters enormously to tax and estate planning.
Consider a sophisticated estate-planning client, one with multiple operating businesses, partnership and LLC interests, S corporation stock, substantial retirement accounts, life insurance, existing irrevocable trusts, prior taxable gifts, family limited partnerships or LLCs, charitable interests, real estate, outstanding debt, complex estate-planning documents, recent appraisals, tax returns, shareholder agreements, and an anticipated business sale.
Historically, a lawyer and several members of a legal team might spend days or weeks simply organizing this information. AI can increasingly assist with that initial process. A properly configured system can help identify who owns what, directly and indirectly; the client’s basis, carryovers, prior gifts, exclusions, and other relevant tax attributes; which trusts own which entities and which entities own which assets; what is currently included in the client’s estate and what could become included in the future; whether the operating agreement matches the estate plan and whether the trust agreement matches the client’s stated objectives; and obvious planning opportunities involving lifetime gifting, GST planning, §1202, §6166, charitable planning, insurance, valuation, or business succession.
The lawyer still has to determine whether those observations are correct. But the lawyer no longer necessarily needs to spend the first several days simply discovering the information. That is a profound change.
The easiest work for AI to absorb is work that is repetitive, rules-based, document-intensive, capable of being checked against authoritative sources, and relatively predictable. That includes much of initial legal research, document comparison, first drafts of memoranda, correspondence, due diligence, tax-return review, entity charts, trust summaries, drafting checklists, issue spotting, document summaries, discovery and document review, preliminary tax calculations, regulatory monitoring, comparison of statutory provisions, first drafts of trust provisions, and routine client communications. That does not mean these tasks will disappear overnight. It means that the amount of human time required to perform them will decline dramatically, and that has major economic consequences.
II. Where Judgment Still Matters Most
Take a relatively straightforward estate plan. A client needs a revocable trust, pour-over wills, financial powers of attorney, health care documents, beneficiary designations, and perhaps an irrevocable life insurance trust. AI will increasingly be capable of producing the first draft of virtually all of those documents, so the lawyer’s traditional value proposition, “I can draft these documents for you”, will become less compelling.
Now consider a different client: one who owns a $150 million business and is considering selling it. There are three children, one works in the business, two do not. The client has made $20 million of prior gifts, has existing dynasty trusts, and has substantial life insurance. The business may qualify for §1202. The client is concerned about control, the client’s spouse has different objectives, and the client wants to provide for grandchildren without giving children outright control.
Now the question is no longer “Can you draft my trust?” The question is “What should I do?” That question is much harder to automate.
The most valuable estate-planning attorney of the future will not necessarily be the person who can draft a trust fastest. The attorney will be the person who can look at the entire situation and say: “Here are the three structures I think you should consider. Here is what each accomplishes. Here are the tax consequences. Here is what you give up. Here is what happens if the business sells for $100 million, $250 million, or $500 million. Here is how your children are affected. Here is how control changes. Here is what happens if you die tomorrow. Here is what happens if you live another 30 years. And here is the structure I recommend.” That is strategic judgment. AI can assist tremendously with that analysis, but the lawyer remains responsible for the recommendation.
III. How the Partner’s and Associate’s Roles Will Change
For decades, the traditional law-firm model has run from partner to associate to research to draft to review to client. AI begins to compress that pyramid. The future may look more like client information flowing to AI analysis, then to partner judgment, then to strategy and implementation. The partner may personally perform substantially less mechanical work while becoming more deeply involved in the most important decisions.
This has a major implication: partners will increasingly be paid for judgment rather than production. A partner who can use AI to accomplish in two hours what once required ten hours of associate research should not conclude that the legal service has become worthless. Instead, the partner should ask what additional value can be provided with the eight hours now freed up. The answer may be better planning, more alternatives, deeper client counseling, more frequent client contact, better business understanding, more sophisticated transaction structures, stronger relationships with CPAs and financial advisors, greater business development, and more time spent identifying opportunities for the client. That is where the future economic value lies.
The most disruptive effect of AI may not initially be on senior lawyers — it may be on associates. For generations, associates learned by performing assignments such as researching an issue, reading cases, reviewing an agreement, drafting a memo, preparing a trust provision, or comparing documents. AI can increasingly do the first draft of all of those assignments. That creates a difficult question: how does a young lawyer become an expert if the machine is doing the work through which the lawyer traditionally acquired expertise? This is not a theoretical problem. Legal leaders are already discussing the effect of AI on junior lawyers and whether traditional training models will survive, and even AI-industry leaders acknowledge the concern that junior lawyers may lose some of the traditional work through which they developed legal judgment. Law schools are wrestling with the same problem; some are limiting AI use in foundational courses precisely because they are concerned that students can become dependent on technology before developing independent analytical skills.
Young lawyers should take AI seriously. The associate who believes that legal research, drafting, and document review will remain essentially unchanged for the next 20 years is taking a significant career risk. But the answer is not to compete with AI, it is to develop capabilities that AI cannot easily replace. An associate should become exceptionally good at understanding the facts, since the best tax analysis is worthless if the lawyer does not understand what actually happened; understanding the business, since a tax lawyer who understands how the client’s company makes money will outperform one who merely understands the Internal Revenue Code; exercising judgment, because AI can generate five possible answers but the lawyer has to decide which one is actually right for the client; communicating clearly, because clients don’t want a 40-page AI-generated memo, they want to know what the lawyer recommends and why; and developing relationships, because the associate who knows the client, the CPA, the investment banker, the valuation expert, and the client’s business objectives is building something AI cannot simply download.
There is also a potentially positive side to this transformation. If AI eliminates large amounts of low-level production work, a talented associate may be able to move much faster into sophisticated analysis, using AI to understand a new area of law, test multiple planning structures, analyze dozens of authorities, review an enormous client file, identify issues, prepare alternative strategies, have the partner challenge the analysis, and participate directly in the client meeting. Instead of spending five years primarily producing research memos, the associate could begin learning how partners think much earlier. That could accelerate professional development, but only if firms deliberately redesign training.
IV. Billing and Ethics Will Have to Adapt
AI also creates a fundamental problem for hourly billing. Suppose an attorney historically spent 20 hours researching and drafting a complex tax memorandum, but, with AI, can perform the same analysis in five hours. Can the lawyer bill 20 hours? The answer is increasingly clear: not simply because that is what the task historically took. ABA Formal Opinion 512 emphasizes that lawyers using generative AI must comply with their obligations concerning competence, confidentiality, supervision, communication, and reasonable fees, and it specifically addresses the relationship between AI-driven efficiency and billing. This could accelerate the movement toward fixed fees, value-based fees, project fees, subscription models, annual planning arrangements, and hybrid hourly/fixed-fee structures. Ironically, AI may force lawyers to explain their value more clearly, which is probably healthy for the profession.
AI does not relieve the lawyer of professional responsibility, it increases it. The ABA’s Formal Opinion 512 makes clear that lawyers remain responsible for competent representation, confidentiality, supervision, communication, accuracy, and reasonable fees when using generative AI. This is especially important in tax and estate planning because the consequences of an error can be enormous: a hallucinated citation in a research memo is bad, but a hallucinated tax rule incorporated into a $100 million estate plan can be catastrophic. Therefore, the future lawyer must know not merely how to use AI, but when not to trust it. That may become one of the most valuable professional skills of all.
There is a counterintuitive possibility here: AI may make basic legal services cheaper while making sophisticated planning more valuable, because the ability to identify sophisticated strategies will increase. Imagine AI identifies a potential §1202 issue, a possible §2036 concern, a GST allocation problem, an opportunity to use a sale to an intentionally defective trust, a valuation issue, a business succession problem, a charitable planning opportunity, and a potential §6166 problem. The lawyer now has more information, but someone still needs to determine which opportunities actually matter. That is where the sophisticated lawyer becomes more valuable.
V. The Future Client Experience
Imagine a client meeting in 2030. Before the meeting, the firm’s AI system has reviewed five years of tax returns, balance sheets, financial statements, entity agreements, trust documents, insurance policies, prior gift tax returns, estate-planning documents, ownership charts, valuations, and transaction documents. The system produces a planning dashboard identifying estate exposure, projected estate tax, business concentration, a potential liquidity event, existing trust assets, potential §1202 exposure, potential GST opportunities, planning deficiencies, and recommended areas for review.
The attorney then sits down with the client. But the attorney does not say, “My system found seven issues.” The attorney says, “I’ve looked at your entire picture. There are three things I think we should do now, two things I would wait on, and one thing I would not do despite what you may have heard from your other advisors.” That is the future: AI is the analytical engine, and the lawyer will recommend the best path.
VI. Building an AI-Enabled Practice
The answer is not simply “learn ChatGPT.” Every partner should certainly become AI-literate, but the more important objective is to build an AI-enabled practice. A partner should begin asking which parts of the practice are repetitive and automate them; what information is repeatedly analyzed and build systems around it; what documents are repeatedly reviewed and build workflows for them; what questions clients repeatedly ask and turn those into knowledge systems; what is unique about how the partner practices and capture that intellectual capital; and what the partner knows that younger lawyers don’t, turning that into institutional knowledge.The goal is to make the partner’s knowledge scalable.
Consider a senior tax lawyer who has spent 25 years developing sophisticated knowledge of estate tax, GST, §1202, partnership taxation, business succession, valuation, trust planning, charitable planning, corporate transactions, and family wealth planning. Historically, much of that knowledge exists inside the lawyer’s head. AI creates the possibility of turning that knowledge into a practice infrastructure, internal precedent libraries, planning checklists, issue-spotting systems, standardized workflows, AI-assisted research protocols, document-generation systems, client intake systems, planning dashboards, and internal knowledge bases. The partner’s knowledge becomes an institutional asset, which may ultimately be one of the greatest competitive advantages of sophisticated firms.
The old competitive advantage was “I know more law.” The new competitive advantage will increasingly be “I know how to combine legal knowledge, AI, business understanding, judgment, and relationships to solve complicated problems faster and better.” That is a very different skill set, and it favors lawyers who are both technologically capable and intellectually sophisticated.
VII. Will AI Replace Tax and Estate Attorneys?
For routine work, AI will replace a meaningful amount of what lawyers currently do, and there is no reason to pretend otherwise. Some basic estate planning will become dramatically cheaper. Some tax research will become nearly instantaneous. Some drafting will become largely automated. Some junior-level work will disappear. Some clients will ask why they should pay hundreds of dollars per hour for work that AI can produce in minutes. Those changes are real.
But the conclusion that “therefore lawyers will disappear” does not follow. A wealthy family does not hire a sophisticated estate-planning lawyer merely because the lawyer can produce a trust document. They hire the lawyer because someone needs to decide what they should do. That question remains extraordinarily valuable.
The successful tax and estate-planning attorney of 2030 may look very different from the attorney of 2020. The lawyer may spend less time researching, drafting, summarizing, organizing, and reviewing, and more time thinking, counseling, designing, negotiating, communicating, originating, and managing relationships. The practice may become less about producing legal documents and more about designing and implementing wealth and business strategies. That is not the death of the profession, it is potentially the elevation of the profession.
Conclusion
Artificial intelligence will not eliminate tax and estate-planning attorneys. It will eliminate the necessity of doing many things the way tax and estate-planning attorneys have traditionally done them. That distinction is critical.
The partner who continues to sell hours of research and drafting will face increasing pressure. The partner who uses AI to eliminate low-value work and then uses the resulting capacity to provide better strategy, better judgment, better client service, and more sophisticated planning may become substantially more valuable.
For associates, the message is equally important: do not build a career around being the person who can perform tasks that AI can perform. Build a career around understanding why the tasks need to be performed, which answer is correct, and what the client should do. For partners, the mandate is even clearer: do not compete with AI. Build a practice that is better because AI exists.
The greatest tax and estate-planning attorneys of the next decade will probably not be those who resist artificial intelligence, nor will they be those who blindly delegate their judgment to it. They will be the lawyers who understand the technology well enough to use it aggressively and understand the practice of law well enough to know when the machine is wrong.
AI will increasingly become the engine behind the practice. The lawyer will remain the architect.
To discuss further, please contact Sebastian Pascu, Chair of KJK’s Estate, Wealth & Succession Planning Practice Group, at 216.736.7294 or SCP@kjk.com.