Websites which strip image metadata may face future copyright liability

In the US District Court of the Southern District of California, Judge Gonzalo P. Curiel is hearing the case of Steven M. Gardner, vs. CafePress Inc. (Case No. 3:13-cv-1108-GPC-JMA).

CafePress.com's website allows users to upload images for printing on items like hats, T-shirts, other clothing, mugs, home decor, etc. Mr. Gardner has alleged CafePress facilitated the storage and sale of his photographs of Alaskan Wildlife, infringing on his copyright, in large part by stripping the metadata, containing his copyright information, from his images. He has claimed that before CafePress disabled access to his images in response to his lawsuit, more than $6,000 in merchandise printed with his images had been sold.

The crux of the case is the Digital Millenium Copyright Act (DMCA), it's “safe harbor provisions,” and the automatic stripping of image metadata, including copyright information which some websites like CafePress perform.

Under the DMCA, Title II, websites are generally exempt from damages, (“safe harbor”) when they unwittingly store and make available for use, copyright infringing images, as long as they meet specific requirements aimed at preventing copyright violations. Among the requirements are the website must “accommodate and not interfere with standard technical measures” which copyright owners use to “identify and protect” their copyright work.

In the case, CafePress moved for summary judgment claiming it hosted and facilitated the sale of copyrighted photos, and was entitled to the DMCA's “safe harbor” protection from damages.
Judge Curiel, in his February order, denied summary judgment for CafePress.

In his discussion of the order, Judge Curiel said, referring to CafePress' “safe harbor” argument,

Here, the Court finds that, at a minimum, Plaintiff [Mr. Gardner] has offered sufficient evidence to create a dispute of material fact as to whether CafePress's deletion of metadata when a photo is uploaded constitutes the failure to accommodate and/or interference with “standard technical measures.” From a logical perspective, metadata appears to be an easy and economical way to attach copyright information to an image. Thus, a sub-issue is whether this use of metadata has been "developed pursuant to a broad consensus of copyright owners and service providers." Accordingly, the Court cannot conclude, as a matter of law, that CafePress has satisfied the prerequisites of § 512(i) [applicable section of the DMCA].

In his statement, Judge Curiel made four critical observations:

1. CafePress acknowledged that when an image is uploaded to their website its metadata is automatically stripped from the image.

2. It's possible that the automatic stripping of an image's metadata may not accommodate, and in fact, may constitute interference with “standard technical measures” to identify and protect images.

3. Logically, the use of metadata is an easy and economical way to attach copyright information to an image.

4. It is possible that there is a broad consensus among copyright owners and service providers concerning the use of metadata for images.

Thus, Judge Curiel has opened the door to the possibility that websites which strip metadata from images uploaded to their website may loose their “safe harbor” protection, and therefore may face copyright liability, in the future, for infringing images uploaded to their website by third parties.

This could be a major step forward to protect image copyright holders from theft and the orphaning of their work. Many of the images on the Internet become orphaned because their copyright information has been stripped during uploading of the images, eliminating the ability for anyone to determine who owns the image's copyright, and how to contact the copyright holder.

Mr. Gardner will now have the opportunity to prove CafePress isn't entitled to use the “safe harbor” provisions of the DMCA, and are therefore liable for actual damages, due to copyright infringement. (Judge Curiel has given CafePress a partial summary judgment on Mr. Gardner's entitlement to statutory damages and attorney fees, as he ruled Mr. Gardner didn't register his copyright of the images before the alleged infringement began.)

In my opinion, this will hinge on the three parts of the definition of “standard technical measures” to identify and protect copyrighted work, in the DMCA, which basically are:

1. The measures used must be by broad consensus of copyright holders and service providers.

2. The measures are available to any person, reasonably and without discrimination.

3. The measures don't impose substantial costs on service providers, their systems and networks.

CafePress disputes that the use of metadata to identify and protect image copyright holders was “developed pursuant to a broad consensus of copyright owners and service providers in an open, fair, voluntary, multi-industry standards process.”

That assertion doesn't seem to make much sense. Since photography first came into being, photographers have signed, stamped, and dated their prints and added descriptions to them. We can even see this on daguerreotypes from the 1800's. Digital metadata is merely the extension of that concept to digital images. We know metadata is a standard for photographers and users alike from its widespread use. Adobe, for example, includes this data and a way to edit it in their industry standard Photoshop and Lightroom programs.

Every digital camera manufacturer has programmed their cameras to automatically insert metadata into every image they save, from professional DSLR's to the most inexpensive consumer “point and shoot” digital cameras. Even smartphones insert metadata into the digital images they produce. The International Press Telecommunications Council (IPTC) has created a precise metadata standard for the press, and for that matter, anyone's use.

While some major provider websites still strip metadata from images like CafePress, such as Facebook, Instagram, and Twitter, many do not, including Dropbox, Pinterest, Tumblr and Google, plus other providers make excellent use of the metadata in images, such as Smugmug and other online galleries.

Use of metadata is easily achievable. It can be inserted into images by any copyright holder through a variety of inexpensive software programs currently available, and directly through digital cameras, and is read and displayed by countless software programs and websites.

Not stripping metadata clearly doesn't impose substantial costs on any service provider. In fact, they need to proactively program stripping the metadata out of their uploaded images. Websites can merely choose to not strip the data from the image with little cost. Moreover, the amount of storage space the data takes is so small as there is no burden for storing it either.

As an experiment, I stripped a series of images of metadata, and compared their file size. My images contain substantial IPTC metadata entries, including copyright information. No image was reduced in size more than 2KB by removing its metadata, an insignificant amount of storage.

The decision is a positive one for photographers, but it's not even a final determination by this court, and as such, hasn't been tested in other cases in other courts. That said, this could be the first sign, the door has been opened, which will give photographers a real opportunity to identify and protect their images, to prevent image theft, to prevent copyright theft, and to prevent the orphaning of millions of images daily.

Perhaps this decision will, in the future, finally push all the major providers to stop facilitating image and copyright theft by their users by their unwarranted, self-serving automatic image metadata stripping.