Every US Jurisdiction Now Accepts a Digital Seal. Your Firm Still Prints.

Every state board in the country will accept a PDF you never printed. Most firms print anyway.

That gap is the story. For years the honest answer to "can we go all-digital on sealed deliverables?" was "depends on the jurisdiction, and you'd better check." That answer is obsolete. All 51 US jurisdictions — 50 states plus DC — permit digital or electronic seals in some form. The legal blocker is gone. The plotter is still running.

Two things hold the old workflow in place, and only one of them is a real problem. The first is habit: a sealing ritual unchanged since drafting tables, encoded in QA checklists nobody has reopened in a decade. The second is a genuine trust gap in what happens to a sealed PDF after it leaves your office. That one deserves respect. It is where digital sealing actually breaks.

The legal question was settled while nobody was watching

Boards did not flip to electronic seals in one dramatic moment. They amended rules quietly, one jurisdiction at a time, usually by adding a sentence permitting an electronic seal provided the signature stays under the licensee's control and the document is protected against alteration. That is what you get from sixty-nine independent, governor-appointed licensing boards, as NCEES's overview of engineering licensure describes them — no central switch to flip. The result is uniform on the yes/no question and non-uniform on everything else.

Look at the geometry alone. Colorado specifies a circular seal 1.5 to 2 inches under Colo. Rev. Stat. § 12-25-217. Ohio specifies the same range under Ohio Rev. Code § 4733.14. California is tighter — 1.5 inches diameter, Cal. Code Regs. tit. 16, § 411. New York wants 1.75 inches under 8 NYCRR § 68.13. Texas allows 1.5 to 2 inches under 22 TAC § 137.31. Across the 51 jurisdictions there are ten distinct diameter specifications, and every one of those jurisdictions permits electronic seals.

That is the modern compliance problem: not "may I seal electronically" but "does this seal match this jurisdiction's dimensions and content." Carry licenses in five states and you are managing five artifacts whose differences are small enough to miss on a squinted review. Our breakdown of PE stamp size requirements and the state-by-state stamp requirements post exist because that variation is the part people get wrong — not the legality.

What "we still print" really means

Walk the print-sign-scan loop honestly. Export a PDF from the CAD or BIM package. Plot it. Apply a physical stamp and sign across it in ink. Scan the sheets back to PDF. Email the scan.

Every step destroys information. The scan is a raster image — vector text gone, layers gone, searchability gone, a file that is larger and less legible. The ink signature proves nothing cryptographically; it proves someone with a pen was in a room. And the scanned output is trivially forgeable by anyone with a decent image editor. Our piece on document fraud and professional seals covers why a scanned seal is the easiest artifact in this industry to counterfeit convincingly.

The print loop feels safer because it is tangible, and official guidance still reinforces that feeling: the Washington board's page on professional stamps and seals is written entirely around the physical article — what the rubber stamp must show, and the requirement to sign in permanent ink across the face of the seal. It is not safer. It is a chain of custody you can hold, wrapped around a document whose integrity you cannot verify.

The pasted-image seal is the worst of both worlds

Here is the failure mode I see most, and it is more dangerous than printing: the licensee keeps a JPG or PNG of their seal, drops it onto the PDF, types or scribbles a signature beside it, and sends the file.

Understand what that document is. It carries the full legal weight of a seal — you have made a professional representation, your license number is on the page, and you own everything downstream of it. It carries none of the protection. Nothing binds the seal to the page content. Nothing detects a changed dimension, a swapped detail, a revised note. Anyone can extract that image in thirty seconds and apply it to a drawing you have never seen.

You have accepted the liability of a seal and purchased zero tamper-evidence with it. A wet stamp on paper at least requires physical presence to reproduce well. A floating PNG requires a right-click. If that describes your process, the fix is not "go back to printing." It is to understand the difference between an electronic signature and a digital signature, and between the seal and the stamp as concepts. A properly applied digital seal binds a certificate to the document's contents, so any modification after sealing invalidates the signature. That invalidation is the entire product. Without it you have a picture.

How a real digital seal breaks downstream

This is where the sceptics have a point, and their case deserves stating at full strength. You seal a set correctly. Then:

• The plan reviewer prints it. Your signature becomes toner. Verification evaporates at the counter and markup happens on paper.
• A consultant flattens the file. Someone opens your sealed PDF in a markup tool, adds comments, saves a flattened copy to "clean it up." The signature is invalid or stripped, and the output looks identical to the naked eye.
• The set gets recombined. Your sheets are merged into a 400-page submission with a dozen other disciplines. The merge rewrites the PDF structure. Every signature in the package dies at once.
• Gateways and portals rewrite attachments. Some re-encode PDFs for scanning or preview. Some regenerate a "web-friendly" version and serve that as the record copy.
• Someone screenshots the sealed page into a new document — back to the pasted-image problem, except the pasted seal now came off a legitimately sealed sheet, which makes it look more credible, not less.

None of this is exotic. It is the ordinary life of a construction document, and it produces a reasonable conclusion: if the third party destroys the signature anyway, why bother?

The strongest objection, and why it still loses

The objection goes like this. Digital sealing pays off only if the whole downstream chain preserves it. We control none of that chain. So we are buying integrity that gets discarded at the first handoff.

Coherent, and wrong in a specific way: it treats the signature as valuable only at the moment of final consumption. The signature's most important job happens earlier and closer to home.

Consider what you can prove when a dispute arrives two years later. With a digital seal you hold a file whose signature either validates against the exact bytes you issued or does not. That is a mechanical answer to "is this the document I sealed?" — the one question that matters when someone shows up holding a set with your name on it and a detail you would never have approved. With a scan or a pasted image you have no mechanism. You have your memory and your file server.

Second, the chain improves from the far end, not the near end. Plan reviewers who print are not a permanent condition, and every firm that keeps feeding the paper loop is a reason for the reviewer not to change.

Third, and most practically: an invalid signature is information. When a flattened file fails verification, the system is correctly reporting that the document was altered after sealing. That is not the tool failing. That is the tool telling you your set was modified downstream, which is exactly what you want to know. A scanned seal never tells you anything. That is not the same as never being wrong; it is being silently wrong.

This is a process migration, not a software purchase

The common way this fails: a firm buys a tool, tells everyone to use it, changes nothing else. Six months later half the staff are still plotting, half are pasting images, and the tool gets blamed.

1. Fix the artifact first. Get a seal file that matches each jurisdiction you practice in — correct diameter, correct content, correct format. Boring, and the foundation. If you are still adding jurisdictions, NCEES's Records program is the mechanism for carrying transcripts, exam results and references into a comity application rather than re-submitting them to each board. The California engineer stamp and Texas engineer stamp pages lay the differences out plainly, and the free stamp preview tool renders one without a signup so you can check the geometry before committing to anything.
2. Define the record copy. Write down in one sentence which file is the sealed document of record and where it lives. Most integrity disputes are really ambiguity about which of six similarly named PDFs was the real one.
3. Ban the loose image. No seal graphic sits in a shared folder or on a personal desktop where it can be dragged onto anything. Highest-leverage policy change available, and it costs nothing.
4. Instrument the handoff. When you transmit a sealed set, transmit the means to verify it. If the recipient can confirm the document is unaltered without calling you, the signature survived the handoff in the only sense that matters.
5. Handle the print-only reviewer explicitly. Some agencies want paper. Fine — plot from the sealed file, keep the sealed file as the record, and never let paper round-trip back in as a scan that becomes the new master.

Where the process meets the mechanics

Do you still need a signature alongside the seal?

Generally yes. In most jurisdictions the seal alone is not sufficient, and the relationship between the two is worth getting right rather than guessing at — we work through it in stamp and signature: do you need both. In a digital workflow the signature is the cryptographic operation, not a picture of your handwriting, and that distinction matters more than firms expect.

What actually goes on the sheet?

Placement, sizing against sheet scale, staying legible after export — that is the part that trips up people migrating off a physical stamp, and applying a PE stamp to PDF drawings walks the steps. If you are earlier than that and still nailing down what your board expects a seal to contain, start with PE stamp requirements instead of replicating your ink stamp pixel-for-pixel, and read the electronic versus wet stamp comparison before you assume the two are interchangeable.

Architects and surveyors face the same structure with different specifics. The architect stamp and land surveyor stamp requirements diverge from the engineering ones in content, not in the legal availability of electronic sealing.

The obligation the print loop hides

One more reason to stop treating this as optional. When you seal, you are asserting that you sealed it. A process where your seal exists as a portable image, applied by whoever has the file, quietly transfers that assertion to your office at large. Nobody decided that. It happened because the artifact was a picture, and pictures travel.

Digital sealing puts the assertion back where the license sits, because the signing operation requires the credential, not just the graphic. That is a governance improvement independent of any reviewer's habits.

Do this in the next hour

Pick your most recent sealed deliverable. Open the file that was actually transmitted — not the one in your project folder, the one that went out.

Ask three questions. Is the seal a raster image or a signed object? If someone changed a dimension on sheet S-4 tomorrow, what would tell you? And does the seal geometry match the jurisdiction it was issued for, or the state where you first got licensed?

If any answer is uncomfortable, you have found your starting point, and it is a smaller job than a workflow overhaul. Confirm the seal specification for your jurisdiction, check that your license record reads the way you expect through a license lookup, and get one compliant seal file generated correctly before you touch how the firm transmits anything. The artifact is upstream of the process. Fix it first, then go argue with the plan reviewer.