21 December 2025

Recently I was asked by a rich font company to endorse a letter they wanted to send to the U.S. Copy­right Office about “Restoring Copy­right Regis­tra­tion for Font Soft­ware”. The impetus was a set of concerns arising from ongoing rejec­tions of font-copy­right regis­tra­tions and a recent opinion in a case called Laatz v. Zazzle pertaining to the infringe­ment of font copy­rights.

I didn’t add my name to the letter. For several reasons. First: I avoid doing free work for rich compa­nies whose inter­ests only super­fi­cially overlap with mine. Second: I’ve never regis­tered a copy­right in my fonts, so the rele­vance seemed faint. Third: digital fonts (prob­ably) aren’t protected by copy­right, so the whole premise of the effort seemed fatally flawed.

Please note: I’m a copy­right liti­gator and a type designer. Still, the surrounding discus­sion is not legal advice. Rather, it’s a summary of how I reason about certain issues in U.S. copy­right law that affect me as a partic­i­pant in the font industry. Some of my views may be widely shared; some may be shared by no one; some may turn out to be completely wrong. I treat the terms font and type­face as func­tional synonyms.

Further note to fellow type designers: I love you all, but consis­tent with this prin­ciple, I won’t be able to respond to private messages about this piece. To avoid conflicts of interest, I’ve never repre­sented other type designers in legal matters. If you have ques­tions about the legal status of your own work, contact your lawyer, not me.

The copyright status of digital fonts: doubtful

Although there has never been a U.S. court deci­sion directly ruling on whether digital fonts are protected by copy­right, it seems reason­able to conclude that the answer is no, based on the following author­i­ties:

Fonts have tradi­tion­ally been excluded from U.S. copy­right protec­tion. This prin­ciple was judi­cially affirmed in the 1978 case Eltra Corp. v. Ringer (“type­face has never been consid­ered enti­tled to copy­right”). 1.

After digital fonts arose, a 1988 USCO regu­la­tion confirmed the vitality of this general rule (“digi­tized repre­sen­ta­tions of type­face designs are not regis­trable under the Copy­right Act”). But this regu­la­tion still permitted regis­tra­tion of certain soft­ware programs that operate on digital fonts. 1.

A 1992 USCO regu­la­tion amended the 1988 regu­la­tion by updating regis­tra­tion require­ments for font-related soft­ware programs. (Around this time, a sepa­rate USCO regu­la­tion about “Mate­rial not subject to copy­right” was also updated to list “Type­face as type­face.”) 1.

A 1998 federal district-court opinion in Adobe v. Southern Soft­ware found that certain font soft­ware programs were enti­tled to limited copy­right protec­tion because there was “some creativity … [in] the selec­tion of the points.” Since then, some type designers have read this opinion to mean that—contra Eltra and the 1988 USCO regu­la­tion—digital fonts are copy­rightable, if presented to the USCO in the guise of soft­ware programs. 1.

The 2021 revi­sion of the Compendium of U.S. Copy­right Office Prac­tices affirmed the long­standing USCO posi­tion that “The copy­right law does not protect type­face”. 1.

A 2024 federal district-court opinion in Laatz v. Zazzle was appar­ently the first to analyze the Adobe deci­sion. Revis­iting its facts, the Laatz court suggested that the applic­a­bility of the Adobe deci­sion is more limited than what type-designer urban legend might suggest.

Consis­tent with the above, I think that if a court were asked to directly consider the copy­righta­bility of an ordi­nary digital font, it would likely rule in the nega­tive.

Registering fonts as software programs

Some type designers have inter­preted Adobe v. Southern Soft­ware to mean that the place­ment of the points that define each glyph outline consti­tutes copy­rightable expres­sion. Thus, the thinking goes, as long as you repackage a font in the guise of a soft­ware program—say, by auto­mat­i­cally converting it to XML—you can register it as a soft­ware program that gener­ates a font (which is poten­tially regis­trable) rather than as a font per se (which is not). Over the years, some type designers have reported getting font copy­right regis­tra­tions this way. Which in turn fueled the urban legend.

Even assuming this is a correct reading of Adobe—there is increasing reason to suppose it neither is nor was—there are still two prob­lems with this supposed workaround:

A font converted auto­mat­i­cally to some­thing that looks like a soft­ware program may sepa­rately violate the human-author­ship require­ment. Copy­right protects human-made works, not machine-made. For that reason, works presented to the USCO for regis­tra­tion must be authored by a human, including programs that generate a type­face. (The human-author­ship require­ment is also why the USCO has refused to register copy­rights in entirely AI-gener­ated works.) 1.

Even assuming a font-gener­ating soft­ware program qual­i­fies as human-authored, the USCO will still exclude the gener­ated font from the scope of the copy­right (“the regis­tra­tion does not extend to any type­face … that may be gener­ated by the program”). In other words, even if you manage to get a copy­right regis­tra­tion via this contorted method, the scope of that regis­tra­tion still won’t cover the font itself, defeating the point of the contor­tion.

Registration does not confer validity

“But the USCO regis­tered my font based on my converting it to XML [or some other soft­ware-like nota­tion], there­fore my font has a valid copy­right.” Leaving aside the issues noted above, this kind of finders–keepers reasoning is legally unsound.

Under U.S. law, an author gains a copy­right at the moment an eligible work is created. Regis­tra­tion of the copy­right with the USCO is a sepa­rate, largely optional step. Regis­tra­tion only becomes manda­tory if one wants to sue for infringe­ment of that copy­right. (Early regis­tra­tion of a copy­right also improves the legal reme­dies avail­able for a future infringe­ment action, including statu­tory damages.)

But even if a work is success­fully regis­tered with the USCO, it is no guar­antee of the under­lying legal validity of the copy­right so asserted. The regis­tra­tion is merely a confir­ma­tion that the USCO has inspected your appli­ca­tion and concluded that you’ve met the regis­tra­tion criteria. The USCO might be wrong.

For that reason, if a copy­right regis­tra­tion becomes the basis of a copy­right-infringe­ment lawsuit, the defen­dant can still chal­lenge the validity of the copy­right—say, by offering evidence of misrep­re­sen­ta­tions on the regis­tra­tion appli­ca­tion, or showing that the work was never eligible for copy­right protec­tion in the first place.

Font copyrightability after Laatz v. Zazzle

This issue of font-regis­tra­tion validity surfaced in Laatz v. Zazzle, a case filed in 2022. Type designer Nicky Laatz alleged that Zazzle used certain Laatz fonts without permis­sion. (I have no connec­tion to either party.) Laatz sued Zazzle for copy­right infringe­ment and other related legal claims.

In March 2024, the Laatz court issued a ruling on the prelim­i­nary validity of Zazzle’s asser­tion that Laatz’s font copy­rights were invalid. Laatz had defended the validity of her font-copy­right regis­tra­tions by citing Adobe v. Southern Soft­ware. Thus, the Laatz court became the first (that I know of) to judi­cially examine the reasoning of Adobe.

Font-copy­right nerds can read the rele­vant part of the opinion and make up their own minds. In revis­iting Adobe, the Laatz court points to details that distin­guish the facts of that case from those of Laatz. In so doing, the Laatz court suggests that the legal conclu­sion of Adobe doesn’t neces­sarily apply beyond its partic­ular factual context:

As an initial matter, the Laatz court observes that “Laatz used [commer­cial type-design soft­ware program] FontLab to draw and/or select the coor­di­nates to create a digital repre­sen­ta­tion of each type­face, and the FontLab soft­ware gener­ated the code contained in the computer files Laatz regis­tered.” 1.

The Laatz court then contrasts these facts with those of the Adobe case: “Adobe purchased digital font files and trans­lated them into its own font coor­di­nate system using internal Adobe soft­ware … Notably, and unlike this case, the copy­right holder in Adobe created the soft­ware that produced the font programs.” (emphasis added) 1.

Because of this factual distinc­tion, the Laatz court concludes that the judge in the Adobe case “had no occa­sion to consider whether a work is copy­rightable where, as here, the copy­right holder merely ‘selected coor­di­nates and various other points of instruc­tion’ and a third-party soft­ware wrote the imple­menting code. … This distinc­tion is impor­tant … Adobe does not support Laatz’s argu­ment that her copy­right regis­tra­tions are valid.” (emphasis added)

To those in the type industry who have staked a lot on the Adobe case, that last sentence might be a doozy. The Laatz court’s perspec­tive debunks decades of wishful thinking about the breadth of the Adobe opinion. Under the Laatz view, unless you “created the soft­ware that produced the font programs”, you don’t fall within the scope of the Adobe ruling.

(PS—the Adobe and Laatz deci­sions arise from federal district courts and thus don’t have formal prece­den­tial value as caselaw: other federal judges are not required to follow their reasoning. Even so, as the only opin­ions studying this issue, they are likely to remain infor­mally persua­sive for other courts.)

Mean­while, the Laatz case continues. In November 2024, Zazzle asked the court to dismiss Laatz’s copy­right-infringe­ment claim because evidence devel­oped during the case had shown that her font copy­rights were invalid. In August 2025, the court did so. But to reach that conclu­sion, the Laatz court focused on flaws revealed in how Laatz prepared her specific regis­tra­tion appli­ca­tions, and did not consider further the larger issue of copy­righta­bility of fonts.

The lurking horror

At least one rich font company has suggested to me that the Laatz opinion—and the USCO regis­tra­tion prac­tices revealed therein—signal a change since the Adobe deci­sion in the nature of copy­right protec­tion in fonts.

I disagree. I think the Laatz court showed how long­standing USCO regu­la­tions harmo­nize with the Adobe case. Nothing in Laatz or Adobe is incon­sis­tent with the general prin­ciple (since at least 1909) that fonts are not copy­rightable, nor the excep­tion (since at least 1988) that certain font-gener­ating soft­ware programs might be.

To my mind, the ques­tion that flows most natu­rally from Laatz is different: how many existing regis­tra­tions of font copy­rights might be vulner­able to a similar attack during liti­ga­tion? As noted above, for a long time, type designers passed around the folk wisdom that converting a font to some soft­ware-like format (say, XML) was suffi­cient to register a font as a soft­ware program. I’m sure some of these were accepted by the USCO. The Laatz case offers a cautionary tale: it shows how a designer can start an infringe­ment case with appar­ently valid font copy­rights, but end the case without them, having been deto­nated during liti­ga­tion by the defen­dants. Need­less to say—a copy­right regis­tra­tion that’s unlikely to survive infringe­ment liti­ga­tion isn’t worth much.

Why don’t I register my font copyrights?

Partially because I’ve never seen enough evidence that digital fonts are copy­rightable. There­fore, the urban legends suggesting certain workarounds to register fonts with the USCO—e.g., converting a font to some­thing that looks like a soft­ware program—are a waste of time because regis­tra­tions so achieved are unlikely to stand up in court. (Though iron­i­cally, as the rare type designer who “created the soft­ware that produced [his] font programs”, perhaps I would be eligible for the Adobe excep­tion.)

But more impor­tantly, in prac­tical terms—what would be the point? Since 2011, I’ve run a small font busi­ness. Not long after I release a font, it will be uploaded to some public pirate-soft­ware website. I can’t control that. Like every other kind of digital-media file, anyone who wants to pirate my fonts can do so if suffi­ciently moti­vated.

For that reason—and inde­pen­dent of copy­right law—my busi­ness neces­sarily runs on some­thing more akin to the honor system. I try to make nice fonts, price my licenses fairly, and thereby make internet strangers enthu­si­astic about sending me money rather than going to pirate websites. Enough of them do. My busi­ness continues. (Indeed, in terms of rational economic choice, I’ve argued that soft­ware piracy doesn’t exist.)

As an inde­pen­dent type designer, my busi­ness is different from that of a rich font company with zillions of legacy fami­lies. The long-term value of that busi­ness prob­ably depends largely on that company’s ability to assert and defend its intel­lec­tual prop­erty, and extract ongoing rents there­from. My busi­ness, by contrast, depends more on attracting new customers and releasing new work. It’s like comparing an oil tanker to a lobsterman. Both busi­nesses involve boats. Beyond that, our inter­ests diverge.