Why Legal Entity Management is More Than a Register

A director’s resignation from a holding company and two subsidiaries can trigger different filing requirements in every jurisdiction where those entities are incorporated. In one jurisdiction that means a filing within 14 days. In another it means a board resolution, a certified translation, and a form the company secretary has not touched since last year. None of it is difficult on its own. But all of it has to be right, and it has to be right in five places at once.

Now consider a group with 40 subsidiaries across a dozen jurisdictions that runs this loop constantly, through director changes, share transfers, annual returns, and the occasional restructuring. The register shows what exists and when the next filing falls due. It cannot read the articles of a Spanish subsidiary and tell you how this year's annual return differs from the German subsidiary down the list. Someone has to open the documents, compare them, and reason through what each jurisdiction wants. That reading is the quiet, recurring cost of running a multi-entity group.

Miss a step and the cost lands quickly. An entity can lose good standing, which can stall a financing or a sale until it is cured. Directors and officers can carry personal exposure for late or inaccurate filings. A deal team can ask for a clean picture of the group on a Friday and find that three subsidiaries are mid-correction. The work of keeping every entity current and consistent is unglamorous, and it protects the whole organization from problems that surface at the worst possible moment. This article is about the gap between the register and the reasoning behind it, and how in-house teams are closing it.

Legal Entity Management in Plain Terms

Legal entity management is the ongoing work of keeping a company's legal entities in good standing. It means maintaining accurate records of directors, officers, shareholders, and governing documents for each entity, and meeting the statutory filing obligations in every jurisdiction where the group operates.

In practice, that work runs on two tracks. The first is the record itself: the current list of who sits on each board, who holds the shares, which entity owns which, and when each filing is due. The second is the set of governing documents behind every entry: the articles of association, bylaws, constitutions, shareholder agreements, and past filings that define how each entity is actually run.

Keeping the record current is demanding but well understood. It means tracking statutory deadlines, filing annual returns on time, updating director and officer details after every change, and maintaining the beneficial-ownership registers that a growing number of jurisdictions now require. Company administrators and corporate-services teams own most of this, working alongside legal operations and local counsel in each jurisdiction where the group holds entities.

The discipline sounds procedural, and much of it is. The judgment shows up on the second track, in reading and reconciling the documents behind the record.

Where the Entity Register Reaches its Limit

An entity management platform holds the entity list, the ownership structure, the document files, and the filing calendar, and it reminds you when something is due. For a group with hundreds of entities, that record is the difference between order and chaos, and no serious legal team should run without one.

Its job ends at the record. The platform can tell you that an annual return is due for a subsidiary next month. It cannot open that subsidiary's articles and tell you what this year's return has to contain in that jurisdiction, or how that differs from the subsidiary incorporated next door. The platform keeps the register current. The legal question sits one layer down, in the governing documents behind each line of that register.

Consider what happens when a rule changes. A jurisdiction introduces a new beneficial-ownership requirement, and the platform can flag that a filing is now due for every entity there. It cannot tell you which of those entities already satisfy the requirement in their existing articles, which need a fresh resolution, and which raise a question the local rule leaves open. That reading is a person's job, and it repeats every time a regime shifts.

This is where the real measure of complexity hides. Teams tend to size an entity portfolio by counting entities. The work, though, scales with the number of legal regimes and languages the portfolio spans, because that is the reasoning surface. A group of 60 subsidiaries sitting in three jurisdictions is far lighter to run than 30 subsidiaries spread across 20. The rows look similar. The reasoning does not.

According to The Accelerating Impact of Legal AI, most in-house teams and firms deploying Harvey's agents are putting them to work on contracting, compliance, and regulatory tasks, the same recurring, rules-driven texture as entity obligations. The pattern is telling. The work moving to legal AI first is the high-volume reasoning that used to sit in a queue, waiting for a lawyer to find time.

Does Multi-Jurisdiction Compliance Scale With Entity Count?

Each subsidiary is its own small legal world. It has its own filing calendar, its own conventions for how articles and constitutions are drafted, its own regulators, and often its own working language. A group with 60 subsidiaries across 15 jurisdictions is really 15 sets of rules to hold in view, refreshed every time a rule changes in any one of them.

That creates a specific kind of reasoning work. Someone has to compare how an annual return differs across jurisdictions and flag where the obligations diverge. Someone has to check that the articles of association and constitutions across subsidiaries still line up where the group needs them to, and note where local law has pulled them apart. And someone has to work with governing documents that arrive in several languages, holding the meaning steady across each translation.

The divergence is rarely dramatic, and that is what makes it costly. One jurisdiction wants an annual return that confirms the current directors and registered office. Another wants updated articles filed whenever the share capital changes. A third treats a late confirmation statement as a criminal matter for the officers. Holding all of that straight across a portfolio, and spotting the one subsidiary where the rule quietly changed this year, is the work that fills a corporate-services team's calendar.

This is the work a reasoning layer can take on. For example, Harvey can run the legal research to compare filing requirements across the jurisdictions a group operates in, and summarize what an annual return calls for in each. It can read governing documents in several languages and hold the analysis consistent across them, so a change in one subsidiary can be checked against its siblings without a separate manual review for each.

Consider Talanx, one of Europe's largest insurance groups, which brings brands such as HDI and Hannover Re together across 175 countries. Its legal team uses Harvey to review 100-page foreign court decisions in minutes and to produce consistent outputs across jurisdictions, with multilingual contracts and translations part of the daily routine. On one recurring review tied to a new regulatory regime, the team cut the work from about two hours to 15 minutes. Over a single year, that discipline saved more than 400 hours of external consultant time.

Across all of this, Harvey produces a first pass that a qualified lawyer reviews before the team relies on it. Harvey reads, compares, and summarizes the documents, and the register remains the group's book of record.

Who Reads the Articles of Association Across Every Subsidiary?

The register points to the documents. The documents hold the answers. A single subsidiary can carry a stack of bylaws, articles, shareholder agreements, and historic filings that no one has read in full in years. The answer to a live question, who can sign what a quorum requires, whether a transfer needs consent, usually sits somewhere inside them.

The reading work has a few recurring shapes. A team needs to turn a subsidiary's governance documents into a short, usable summary. It needs the current director and officer list pulled from corporate registries or the latest filings. It needs legal document comparison across subsidiaries, reading the articles or constitutions to see where they diverge. And it needs dense bylaws turned into plain-language guidance, sometimes in another language so a local team can use it.

Harvey does this reading work directly. It summarizes governance documents into a brief a lawyer can scan, extracts director and officer details from filings, and compares constitutions across subsidiaries to surface the differences that matter. Deutsche Telekom, the telecommunications group operating across many jurisdictions, identified around 50 priority uses for Harvey out of roughly 400 it reviewed, among them legal text analysis, contract comparison, and document translation. Its lawyers describe Harvey as one place to analyze, compare, draft, and translate dense material that once meant opening a dozen files by hand.

Extraction is worth pausing on, because director and officer records drift more than teams expect. People resign, boards reshuffle, and a registry filing lands months after the meeting that prompted it. Pulling the current list straight from the filings gives a team firmer ground than a field someone updated by hand. It is often the difference between a clean group picture and an awkward correction in the middle of a deal, where AI for due diligence can expose a stale record in minutes. The same reading turns a dense set of bylaws into plain-language guidance a business team can act on without a lawyer decoding every clause.

Summarizing and extracting are exactly the kind of output a team will act on, so the review discipline holds here too. Harvey produces the summary, the list, or the comparison, and a qualified lawyer checks it against the source before anyone relies on it. Harvey does the reading and the drafting, and the register and the underlying filings stay the record.

Bringing Corporate Entity Work In-House

For years, the cost of understanding each jurisdiction's rules made outside counsel the default for even routine entity steps. If confirming what a Dutch annual filing requires meant hours of research or a call to a local firm, sending the whole task out was the rational choice. When that research gets faster and cheaper, the make-or-buy line moves, and more of the work comes back in-house.

What a team can produce in-house grows accordingly. First-pass comparisons of filing obligations across jurisdictions, plain-language summaries of governance documents, and drafts of the recurring work product itself all become in-house tasks. A team can draft template shareholder resolutions from its own approved exemplars in Harvey, then route them to a qualified lawyer to review and finalize before they are used.

This does not pull everything in-house, and it should not. Novel local-law questions, contentious matters, and anything that turns on a judgment call a regulator might test still belong with specialist counsel. What changes is the long tail of routine work — the confirmations, comparisons, and first drafts that were sent out mainly because checking them in-house was slow.

The shift shows up in the research. According to The New Economics of In-House Legal, more than a third of in-house teams report bringing work back in-house that they used to send to outside counsel. Most point to reduced external spend as a reason they justify their investment in Harvey, individual users report saving four to eight hours a week, and headcount stays flat even as the work absorbed grows. The economics move quietly, one recurring task at a time.

Repsol, the global energy group, runs a legal division across 12 jurisdictions and reached 96% adoption of Harvey among its lawyers. With that reach, the team has taken corporate matters in-house that it would once have sent outside, doing the research and first drafts internally and reserving outside firms for the judgment calls that warrant them.

None of this removes the lawyer or the filing. Harvey does the reasoning and produces the draft, a qualified lawyer reviews it, and the filing itself stays with the team, the registered agent, or local counsel. The judgment call is still a lawyer's to make.

Deploying a Reasoning Layer Over the Register

The workable model is straightforward. Keep the entity platform or register as the book of record. Add a reasoning layer over it for the document and obligation work. Hold a firm review discipline over everything that reasoning layer produces. Each part has a clear job, and the boundaries between them are what keep the arrangement trustworthy.

Starting is a matter of picking one task and proving it. Choose the recurring, high-volume work that hurts most, annual-return comparison across jurisdictions, say, or the filings that follow a director change, and run it through the reasoning layer. Check the output against the register and the source documents. When it holds up on the task you know best, extend it to the next one.

The boundary is worth stating plainly. The register stays the record. The reasoning layer reads, compares, summarizes, and drafts. A qualified lawyer reviews the output before anyone relies on it. That review reflects a lawyer's duties of competence and supervision, and the guidance in ABA Model Rule 1.1 Comment 8 that competence now includes understanding the benefits and risks of the technology a lawyer uses.

Running this work through one reasoning layer has a second benefit. Because the analysis, the comparisons, and the drafts pass through a single place, leadership gets a consistent and auditable view of how the group's entity obligations are handled across jurisdictions. That visibility is hard to get when the same work is scattered across local firms and inboxes. That single path also feeds the group's legal knowledge management, leaving a record of what was reviewed and when that an auditor or a board committee can rely on.

The Work Behind the Register

An entity register does an essential job. It keeps an accurate list of directors, officers, shareholders, and governing documents, and it tracks when each filing is due. The legal work sits one layer down, in reading, comparing, and drafting against those documents across every jurisdiction the group touches. That reading is what turns a current record into answers a team can act on.

The volume of that work rises with the number of jurisdictions and languages a group operates in, well past what an entity count suggests. Teams that treat the reasoning as the real work handle recurring filing and governance work with more consistency and a clearer view of the whole group. They get there by holding a lawyer's review over everything the analysis produces. The register tells them what they have. The reasoning tells them what to do about it.

Harvey is the reasoning layer that sits over the register. It summarizes governing documents, extracts director and officer details, compares articles and filing requirements across subsidiaries, and drafts the recurring work product for a qualified lawyer to review and finalize. If your organization is carrying a growing portfolio of entities across jurisdictions, book a demo to see how Harvey handles the reasoning behind your register.

Frequently Asked Questions

What is the difference between entity management software and legal entity management?

Entity management software is the record. It stores the entity list, ownership structure, documents, and filing deadlines, and it reminds you when something is due. Legal entity management is the wider discipline that includes the legal reasoning on the documents behind that record, such as comparing filing requirements across jurisdictions and reading the articles behind each entity. The software keeps the register, and the discipline covers the judgment the register cannot.

Can AI file annual returns or maintain the entity register?

No. The filing itself stays with your team, a registered agent, or the relevant registry, and the register remains your book of record. AI can take on the reasoning and drafting around those steps, such as comparing what each jurisdiction's return requires or preparing a first draft. A qualified lawyer reviews that work before anyone relies on it.

How does Harvey handle governing documents in multiple languages?

Harvey can read and summarize governing documents in several languages and keep its analysis consistent across them, so articles in one subsidiary can be compared against siblings drafted in another language. A qualified lawyer should review the summary or comparison before the team relies on it.

Does using AI for entity work create compliance risk?

The risk comes from relying on output without checking it, which is why a qualified lawyer reviews everything the tool produces before it is used. Handled that way, the practice sits comfortably within a lawyer's duties of competence and supervision, including the expectation in ABA Model Rule 1.1 Comment 8 that competent lawyers understand the technology they use. Verification discipline is what keeps AI-assisted entity work sound.