Regulatory Disclosure Risk From M&A Language Inconsistencies

Most leadership teams file narrative inconsistency under "messaging problem," something the communications team cleans up before the annual report goes out. That's a fine way to think about it when nothing's at stake. Overall SEC comment letter volume declined in the 12-month period ended June 2025, according to a Harvard Law School Forum on Corporate Governance analysis. But the second a company enters an M&A process, that same inconsistency becomes a legal problem instead of a polish issue, because regulators and counterparties now scrutinize it directly. Companies now have to rely on unqualified representations that they have a reasonable basis to exclude a shareholder proposal under Rule 14a-8, since the Staff will issue a no-action response based only on that representation, not on any independent analysis of the arguments presented [1].

Nobody's checking your work during normal operating life. A company can describe its own business one way in an internal strategy deck, another way on an earnings call, and a third way in a regulatory filing, and in ordinary times none of that gets cross-referenced. That kind of gap is basically background noise in corporate life.

M&A flips the switch. The SEC's Division of Corporation Finance reads the earnings call transcripts. It reads the proxy. And it compares them, line by line, looking for daylight between what management said out loud in a Q&A session and what the proxy claims about synergies, deal rationale, or risk. That comparison is a documented, known source of comment letters. So the accumulated, unglamorous drift in how a company talks about itself, built up over years of nobody paying close attention, has a name once it hits this kind of scrutiny: language debt. It's fragmented terminology racked up across product lines, regions, and however many leadership changes, sitting quietly on the books until a transaction forces someone to read it all at once.

What the SEC is reviewing for inconsistencies

Two documents do most of the damage here, and it helps to know exactly which ones. The merger proxy, filed first as a preliminary PREM14A and later as the definitive DEFM14A, generates more comment letter activity at announcement than anything else in the deal packet. And this isn't a "fix it later" kind of comment. Comment letters on the preliminary proxy have to be resolved before the definitive proxy can be filed and the shareholder vote scheduled. An unresolved language gap doesn't just cost you an awkward phone call with outside counsel. It stops the deal clock, full stop.

Inside that proxy, one section draws a disproportionate amount of attention: "background of the merger," required under Item 14 of Schedule 14A. This is the blow-by-blow account of how the deal came together, and both SEC staff and plaintiff firms comb it for omissions, missing negotiation steps, unnamed alternative bidders, or a rationale for accepting one bid over another that doesn't quite add up. Mindbody and Pattern Energy, decided in 2023 and 2024 respectively, set the bar here, and the lesson from both is blunt: more detail beats less, every time.

None of this happens in a vacuum, either. Guidance reinforced by Katten's 2026 proxy season update, published January 20, 2026, specifically discourages generic, boilerplate risk factor language and requires company-specific, tailored disclosures instead. And MD&A quality, along with whether the pro forma financials actually hold together, are flagged as areas where SEC scrutiny is expected to hold steady or climb through 2026.

Why the acquirer bears the risk the target created

Acquirers shoulder the heavier share of the disclosure burden, which catches people off guard. In most transaction structures, the acquirer shoulders the heavier share of the disclosure burden, full stop, regardless of who created the mess. The target's own historical financials, and how forthcoming it is during diligence, directly shape whether the acquirer can put together an S-4 that's actually accurate. So the acquirer inherits a set of problems it didn't create and, in a lot of cases, didn't even know existed until the paperwork started.

Think about what that actually looks like on the ground. The CFO frames "synergies" with one set of assumptions, while the CEO uses a looser, more optimistic version on stage at an investor day. None of that was malicious. It's just what happens when nobody owns the language across the whole company. And every bit of it becomes the acquirer's problem the day the S-4 gets filed.

Sellers have every incentive to make themselves look good. Revenue projections, retention numbers, profitability stories, these get built to sell the deal, and once the ink dries and the buyer finally has full access to the internal systems, the picture often looks different than the pitch. What surfaced most reliably in 2025 due diligence was a narrative gap, not a spreadsheet error. The narrative gap appears when someone sits down and interviews people directly, and the explanations start shifting between conversations. Third-party verification either backed up the story or blew a hole in it, and in a fair number of cases it turned up undisclosed regulatory exposure or conflicts that no document ever would have shown. Documentation tells you what a company wrote down. It doesn't tell you what people actually say when nobody's checked their story against the last person's.

Expanded HSR Requirements and Transaction Rationale Language as a First-Order Antitrust Issue

If the SEC weren't enough, there's now a second regulator asking a version of the same uncomfortable question. On February 10, 2025, the FTC and DOJ rolled out the most sweeping overhaul of HSR premerger notification requirements in decades, and the new rules demand a far more detailed account of transaction rationale, labor market impact, and supply chain risk. White & Case describes these as the most sweeping changes to the HSR Form for US premerger notification filings in 48 years, with the Federal Register confirming the rule took effect that same February 10, 2025 date.

Pause on what "detailed account of transaction rationale" actually requires. It requires a company to explain, specifically and coherently, why it's doing the deal. That's not a communications quirk. That's a company that cannot produce a single, defensible account of its own transaction when a regulator asks for one.

The stakes only go up as deals get bigger. Cravath's analysis, published through the Harvard Law School Forum, points out that scrutiny scales with deal size, and that companies need a strategy for delivering "consistent and convincing messaging" to regulators, the press, and everyone else, starting on day one. Read that carefully and it makes an assumption most companies don't actually satisfy: that consistent language exists somewhere to be delivered in the first place. For a lot of companies, it doesn't. It gets improvised.

How AI deployment inside the organization accelerates the language problem rather than solving it

AI doesn't invent a company's story. It just amplifies whatever story it's fed. Feed it fragmented, inconsistent internal language, across marketing decks, investor materials, product docs, whatever's lying around, and it will turn out fragmented, inconsistent content, except now at a volume and speed no team of humans could match.

That's not a hypothetical, it's a disclosure problem with a direct line to the SEC's desk. AI-generated content built on ungoverned internal language surfaces its own inconsistencies faster and in greater bulk than any human drafting process ever did, and SEC staff reading filings can see those inconsistencies just as clearly as they can see a human-written contradiction. The root cause, more often than not, isn't a bad model. It's bad data: duplicate records, formats that don't match, systems that don't talk to each other, datasets nobody's updated in years, and rushing AI adoption on top of that mess just compounds it across every output the system produces. Garbage in, garbage out was true before AI. AI just does it faster, and files the garbage in triplicate.

Regulators have noticed. White & Case's alert on annual report risk factors found AI-related risk disclosures grew substantially from 2024 into 2025, with more growth expected in 2026, and the SEC is watching both directions: companies overstating their AI capabilities and companies understating the risks, a pattern regulators have started calling "AI washing". Flip the equation, though, and the same technology becomes an asset instead of a liability. AI trained on a single governed, canonical narrative doesn't amplify chaos, it amplifies consistency, since every document it produces draws from the same underlying source instead of whatever version of the story happened to be lying around that week.

Governed Narrative Infrastructure as a Legal Source Document

Strip away all the specific filings and deadlines, and the root problem is simple: there's no single authoritative version of the company's language that every team is required to use. Without a single authoritative source of language that all teams draw from, there's no document to point to when the SEC asks why the proxy describes synergies differently than the earnings call did. Everyone's just... improvising, individually, and hoping it lines up.

A narrative OS solves that structurally rather than editorially. When it comes time to prepare filings, the consistency isn't something someone has to go back and manually reconcile. It's just already there, baked into how the material was written in the first place.

This is exactly the kind of architecture Storied's Narrative OS is built around: language treated as infrastructure that gets maintained and governed on an ongoing basis, not messaging that gets refreshed every time someone needs a new deck. That distinction matters more than it sounds like it should. Messaging that gets "refreshed" is, by definition, being rewritten under whatever pressure exists at that moment, which is precisely how inconsistency creeps back in. Infrastructure doesn't get rewritten on a whim. It gets maintained.

Timing of Narrative Infrastructure Relative to the Transaction's Start

None of the inconsistency in the public record is fixable at the S-4 drafting stage. By the time lawyers are drafting that document, the inconsistencies are already sitting in the public record, in earnings calls, analyst meetings, press releases, and the internal documents that due diligence is about to dig up. You cannot retroactively make a company's language consistent. You can only discover, at the worst possible moment, that it never was.

That's exactly why the advance-preparation argument matters so much for the biggest deals. Cravath's analysis notes that for mega-deals, getting every front, including communications, in order ahead of time is the only reason these transactions can close as fast as they do. The premium sits entirely on preparation before signing, not cleanup after. And even the regulatory environment's most recent shift doesn't change the underlying requirement. Freshfields' analysis shows the SEC's C&DI revisions gave companies more flexibility in how they structure transactions, but flexibility in mechanics doesn't touch the disclosure consistency bar at all. The narrative still has to hold together, however the deal itself is shaped.

Language debt behaves like a balance sheet liability, a financial framing rather than a communications one. It accrues quietly during the years when nobody's checking, through every leadership change, every regional sales team writing its own version of the pitch, every quarter where nobody reconciled the story. And it comes due exactly when a company can least afford it, in the middle of a transaction, under a regulator's microscope, with the deal clock already running.