Author: Washington Journal of Law, Technology & Arts

First AI Art Generator Lawsuit Hits the Courts

/ Washington Journal of Law, Technology & Arts / 2 Comments
Photo by Steve Johnson on Pexels.com

By: HR Fitzmorris

Your social media accounts may have recently been inundated with spookily elegant renderings of your once-familiar friends’ faces. Or, if you’re on a particular side of the internet, you may have seen any number of info-graphics scolding users for contributing to the devaluation of flesh and blood artists’ livelihoods. What you may not have seen is news of the recent class-action lawsuit filed on behalf of artists who are unhappy with technological advances that, in their view, were ‘advanced’ through art theft.

The Complaint

In the first-of-its-kind proposed class action, named plaintiffs allege copyright infringement, asking for damages to the tune of one billion dollars. Specifically, artists allege that the named AI companies downloaded and fed billions of copyrighted images into their AI software to ‘train’ the artificial intelligence software to create its own digital ‘art.’ In addition to damages, the plaintiffs have asked the court to issue an injunction preventing the AI companies from using artists’ work without permission and requiring the companies to seek appropriate licensing in the future.

The Plaintiffs

The named plaintiffs, who will represent the pool of affected artists if the class is certified by the court, are Sarah Andersen, a popular webcomic artist; Kelly McKernan, who specializes in colorful watercolor and acryla gouache paintings; and Karla Ortiz, a professional concept artist with clients such as Wizards of the Coast and Ubisoft.

In a New York Times opinion piece about the appropriation of her art by both the Alt-Right and artificial intelligence art generators, Ms. Andersen stated, “[t]he notion that someone could type my name into a generator and produce an image in my style immediately disturbed me.” She also explains that the appropriation made her “feel violated” by the way the AI stripped her artwork of its personal meaning and of her human mark that she honed and defined through the “complex culmination of [her] education, the comics [she] devoured as a child and the many small choices that make up the sum of [her] life.” Clearly, for these artists, there is more at stake than the threat to their livelihoods.

The Defendants

The plaintiffs named four entities as defendants in the suit: Stability AI Ltd., Stability AI, Inc., Midjourney, Inc., and DeviantArt, Inc. Each of these companies has a hand in creating, hosting, or perpetuating the use of engines that use AI to create art.

The Legal Issues

The Stable Diffusion engine, for example, is described as a “deep learning, text-to-image model” that anyone can use “to generate detailed images conditioned on text descriptions.” In layperson’s terms, users input text (such as an artist’s name or a specific medium) to generate images with those attributes. This is the heart of the issue. In order to do this, the tool (and others like it) must be “trained,” which involves, in the words of Plaintiff Sarah Andersen: 

[B]uil[ding] on collections of images known as “data sets,” from which a detailed map of the data set’s contents, the “model,” is formed by finding the connections among images and between images and words. Images and text are linked in the data set, so the model learns how to associate words with images. It can then make a new image based on the words you type in.

Stable Diffusion was built using a dataset that contained somewhere in the neighborhood of six billion images culled from the internet without regard to intellectual property and copyright laws or creator consent. Additionally, these companies are not building these engines out of the goodness of their hearts, they are making immense revenue. Stability AI, for example, is currently valued at approximately $1 billion.

The suit, which was filed in the Northern District of California, alleges violations of federal as well as state copyright laws, including “direct copyright infringement, vicarious copyright infringement related to forgeries, violations of the Digital Millennium Copyright Act (DMCA), violation of class members’ rights of publicity, breach of contract related to the DeviantArt Terms of Service, and various violations of California’s unfair competition laws.” The crucial argument for the plaintiffs is that “[e]very output image from the system is derived exclusively from the latent images, which are copies of copyrighted images. For these reasons, every hybrid image is necessarily a derivative work.” (emphasis added).

The defendant companies, though, will likely argue that some version of the “fair use doctrine” protects their activity. To prevail, the defendants must prove that their use of the images was sufficiently “transformative”—unlikely to be confused for, or usurp the market for, the original artwork. 

Whatever the court decides, this type of intersection between art and technology will likely remain a hotbed of intellectual and legal debate as artificial intelligence continues to grow in prevalence and accessibility.

Apple AirTags – Stalking made easy in the age of convenience

/ Washington Journal of Law, Technology & Arts / Leave a comment
Photo Credit: Zachary Zylak

By: Kayleigh McNiel

Marketed as a means of locating lost or stolen items, Apple AirTags are a convenient and affordable tool for tracking down your lost keys, misplaced luggage, and even your ex-partner. Weighing less than half an ounce, these small tracking devices fit in the palm of your hand and can be easily hidden inside backpacks, purses, and vehicles without arousing the owner’s suspicion. 

Reports of AirTag stalking began emerging almost immediately upon their release in April of 2021. Apple’s assurances that AirTag’s built-in abuse prevention features would protect against “unwanted tracking” have fallen woefully short of the reality that these $29 devices are increasingly being used to monitor, surveil and stalk women across the country.

The Wrong Tool in the Wrong Hands – Women Are Being Targeted with AirTags

Through an expansive review of 150 police reports involving Apple AirTags from eight law enforcement agencies across the nation, an investigative report by Motherboard confirmed the disturbing truth. One third of the reports were filed by women who received notifications that they were being tracked by someone else’s AirTag. The majority of these cases involved women being stalked by a current or former partner. Of the 150 reports reviewed by Motherboard, less than half involved people using their own AirTags to find their lost or stolen property.   

AirTags pose a significant danger to victims of domestic violence and have been used in at least two grisly murders. In January 2022, Heidi Moon, a 43-year-old mother from Akron, Ohio, was shot and killed by her abusive ex-boyfriend who tracked her movements using an AirTag hidden in the back seat of her car. In June 2022, Andre Smith, a 26-year-old Indianapolis man, died after he was repeatedly run over by his girlfriend after she found him at a bar with another woman by tracking him with an AirTag.

It’s not just domestic violence victims who are in danger. Stories are emerging on social media of women discovering AirTags under their license plate covers or receiving notifications that they are being tracked after traveling in public places. One woman’s viral TikTok describes how she received repeated notifications that an unknown device was tracking her after visiting a Walmart in Texas. Unable to locate the device, she tried unsuccessfully to disable it, and continued receiving notifications even after she turned off the location services and Bluetooth on all of her Apple devices.   

In January 2022, Sports Illustrated Swimsuit model Book Nader discovered that a stranger slipped an Apple AirTag into her coat pocket while she was sitting in a restaurant. The device tracked her location for hours before the built-in safety mechanism triggered a notification sent to her phone. 

One Georgia woman, Anna Mahaney, began receiving the alerts after going to a shopping mall but was unable to locate the tracker. When she tried to disable the device, she received an error message that it was unable to connect to the server. She immediately went to an Apple Store for help and was told that no beep sounded because the owner of the AirTag had apparently tracked her until she got home and then disabled it. 

Apple’s haphazard release of these button-sized trackers, with near complete disregard for the danger they pose to the public, has resulted in a recent federal class action lawsuit filed by two California women who were stalked by men using AirTags. One plaintiff, identified only as Jane Doe, was tracked by her ex-husband who hid an AirTag in their child’s backpack. The other plaintiff, Lauren Hughes, fled her home and moved into a hotel after being stalked and threatened by a man she dated for only three months. After she began receiving notifications that an AirTag was tracking her, Hughes found one in the wheel well of her back tire. 

The plaintiffs in Hughes et al v. Apple, Inc., 3:22-cv-07668, say Apple ignored the warnings from advocates and put the safety of consumers and the general public at risk by “revolutionizing the scope, breadth, and ease of location-based stalking.” 

The Tech Behind the Tags – Insufficient Safety Warnings and a Lack of Prevention

AirTags work by establishing a Bluetooth connection with nearby Apple devices. Once connected, it uses that device’s GPS and internet connection to transmit the AirTag’s location to the iCloud where users can track it via the Find My app. With a vast network of more than 1.8 billion Apple devices worldwide, AirTags can essentially track anyone, anywhere.  

While the accuracy of Bluetooth tracking can vary, newer iPhone devices (models 11 and up) come equipped with ultra-wide broadband technology that allows AirTag owners to use Precision Tracking to get within feet of its location. 

In its initial release in April 2021, Apple included minimal safety measures including alerts that inform iPhone users if someone else’s AirTag had been traveling with them.Additionally, AirTags chime if separated from its owner after three days. 

When someone discovers an AirTag and taps it with their iPhone, it tells them only the information the owner allows. If an AirTag has been separated from its owner for somewhere between eight and twenty-four hours, it begins chirping regularly. By then, the AirTag owner may have already been able to track their target for hours, learning where they live, work, or go to school. The chirp is only about 60 decibels which is the average sound level of a restaurant or office. This sound is easy to muffle especially if the AirTag is hidden under a car license plate or in a wheel well. This quiet alarm is the only automatic protection against stalking Apple can provide to those who do not have an iPhone. 

Apple did eventually release an app that Android users can download to scan for rogue AirTags, but it requires Android users to know about AirTag tracking and then manually scan for the devices. With only 2.4 stars, many complain that it is ineffective and does not provide enough information.  

In response to the wave of criticism and reports of stalking and harassment, Apple has begun to increase these safety measures in piecemeal updates, which so far have failed to resolve the problem. Just three months after its release, Apple shortened the amount of time it takes for AirTags to chime if separated from its owner; from three days to somewhere between eight and twenty-four hours. But it’s easy to register an AirTag, and then disable it before the target begins receiving notifications.

Our Legal Systems Are Not Prepared to Protect Victims From AirTag Stalking.

Our criminal and civil legal systems are painfully slow to respond to the way technology has changed the way we engage with our families and communities and how we experience harm in those relationships. One of the biggest challenges victims face in reporting AirTag stalking is that many police departments and Courts do not even know what AirTags are or how they can be used to harass and stalk women.

In some states, it is not even a crime to monitor someone’s movements with a tracking device like an AirTag without their knowledge or consent. At least 26 states and the District of Columbia have some kind of law prohibiting the tracking of others without their knowledge. While 11 of these states, including Washington, incorporate this into their stalking statutes, nine others (Delaware, Illinois, Michigan, Oregon, Rhode Island, Tennessee, Texas, Utah and Wisconsin) only prohibit the use of location-tracking devices on motor vehicles without the owner’s consent. These state laws do nothing to protect against AirTags being placed in your bag or purse. These laws also don’t protect those who share a vehicle with their abuser, since the other party is also technically the owner of the vehicle. 

Many states are rapidly seeing the need to beef up their laws in response to AirTags. The Attorneys General of both New York and Pennsylvania have issued consumer protection alerts warning people about the dangers of AirTags. But much more needs to be done.

The fact that Apple released this product without considering the disproportionate impact it would have on the safety of women across the globe shows a clear lack of diversity in Apple’s design and manufacturing process. 

Is Amazon’s APEX the Top Option for Patent Rights?

/ Washington Journal of Law, Technology & Arts / Leave a comment
Photo by PhotoMIX Company on Pexels.com

By: Nicholas Lipperd

Are more avenues to resolve patent disputes a good thing? Patent litigation is a process that can easily cost millions of dollars and which lasts years; it is not exactly an option available to every patent holder. Even with the availability of arbitration, options to protect patents remain limited. Amazon has determined that a private patent evaluation program is a good thing, at least for its Amazon Marketplace. After beta-testing for three years under the name “Utility Patent Neutral Evaluation (UPNE),” Amazon formally implemented its Amazon Patent Evaluation Express (“APEX”) system in 2022, which allows sellers to flag possibly infringing products for Amazon to analyze without the use of the judicial patent system. This system advertises cheap, fast, and fair outcomes to sellers on Amazon Marketplace asserting their utility patent rights, yet has drawn criticism for disproportionately one-sided outcomes leading to its use as a retaliatory tool. Does the fact that this cheap, quick process reduces barriers to litigation offset these shortcomings? Should Amazon make changes to its process to achieve more balanced results?

A case brought in Federal Court for patent infringement takes two to four years to adjudicate, not including an additional year if an appeal is sought. Intrinsically tied to this lengthy timeline is the hefty price tag. Though the median cost for patent infringement cases with $1 million-$10 million at risk fell 250% from 2015 -2019, a full patent trial will still average $1.5 million. How does a patent holder without such resources assert the patent’s rights? Arbitration or mediation are cheaper options, at $50,000 on average, but often requires the other side to agree to participate. When the patent owner wants the patent rights asserted within Amazon Marketplace, though, the owner generally has a cheaper and faster option.

Amazon’s APEX program allows patent holders to have their patents examined by a neutral third-party patent examiner, rather than the United States Patent and Trademark Office (“USPTO”). APEX begins with the patent holder submitting a complaint through Amazon’s Brand Registry, providing the Amazon Standard Identification Numbers (ASINs) of the allegedly infringing sellers and upon which claim in which patent the holder believes the ASINs infringe. For each alleged infringer, Amazon sends a notice and allows up to three weeks for a response. Should Amazon receive no response, such products will be automatically delisted, similar to a default judgment. Upon receipt of the response, an evaluator independent of Amazon and each party is assigned to the issue, and each side is required to pay a $4000 fee, refundable to the winner. The patent holder gets three weeks to submit arguments. The sellers then have two weeks to respond, with the patent holder given one week to submit an optional reply. The evaluator then decides within two weeks, making only the determination if the sellers’ products likely infringe on the patent holder’s claim. It is noteworthy that the APEX evaluator does not make any determination on the validity of the claims in the patent at issue. If the evaluator decides in favor of the seller, the product stays on the platform; if not, the products are removed. There is no appeal process from the evaluator’s decision. The entire process takes fewer than three months, and at a price tag of $4000 per party, creates a fiscal barrier of a fraction of the cost of formal patent litigation.

This process is not, though, without its drawbacks. The patent holder wins a disproportionate amount in APEX proceedings, creating incentives to initiate the process without valid claims. Because the evaluator does not look at the validity of the asserted patent, the accused sellers can do nothing but play defense. In legal terms, they are without the affirmative defense of invalidity. They can’t win, they can only hope to survive. Further, the evaluation is not subject to formal rules like the Federal Rules of Civil Procedure or the Federal Rules of Evidence. The evaluators are hired for their expertise in the patent field, not for their investigative skills in the information provided. With no process of verification from Amazon, patent holders are submitting fraudulent information to obtain favorable judgments. With loose evidentiary rules, a low fiscal barrier, and no chance for the patent to be ruled invalid, the incentives all line up for patent holders to abuse this process, especially considering there is no chance for appeal. Should a competitor be cutting significantly into profits, $4000 is a very low risk for a possibly high reward of ejecting your competition from the market. Tortious interference claims stemming from the APEX process are already coming to light. 

Perhaps the most well-known legal spat involving Amazon’s patent evaluation process is the case of Tineco Intelligence Tech. Co. v. Bissell Inc. (W.D. Wash, 2022). Bissell is a US company that sells vacuums, and Tineco is a Chinese company that does the same. When Bissell initiated a UPNE proceeding, Tineco ignored it, leading to the automatic removal of its products. Tineco moved for a ruling in district court that Bissell’s patent claims were invalid and that their products did not infringe. Luckily, perhaps in part because of the sheer volume of business both entities do, Amazon deviated from its set UPNE/APEX process and reinstated Tineco’s listings before the District Court case finished, though U.S. International Trade Commission (“ITC”) proceedings continued. This case and Amazon’s deviation are seen by some as the exception to the rule. Many entities are still using APEX as a hammer to bludgeon competition into settlements and licensing agreements, despite the tortious interference claims that sometimes follow.

Amazon’s APEX has the potential to be the first of many commercial patent dispute programs due to its budget-friendly, expedited decisions. Yet before it can be considered a system after which other businesses should model their systems, it must rebalance and overcome the issues outlined above. Although a large burden is placed on “neutral evaluators” hired by Amazon, these evaluators currently do not review the patent at issue for invalidity. To establish a more balanced approach and to disincentivize misuse of APEX by predatory sellers, invalidity must be considered. Even if such consideration drives up the required fee slightly, the trade-off would be worthwhile to promote fairness in the process. Amazon has three years of beta-testing under its belt with this system and thus has the data available to see where fraud and misuse are most prevalent. A thorough review of this data should lead to the tightening of its evidentiary standards throughout the process. Despite the name inviting such a pun, APEX must not be allowed to thrive as a predatory tool.

While barriers to justice should not be so high that patent holders may not assert their rights, the process should not be so favorable and easy that it inadvertently incentivizes abuse of the process. Through small tweaks, APEX can continue to serve patent holders’ rights without demanding the time and money that large-scale patent litigation requires.

Virtual Experiences in the Art World: Potential for Copyright Issues

/ Washington Journal of Law, Technology & Arts / Leave a comment
Photo by Erik Mclean on Pexels.com

By: Lauren Liu

Since the COVID pandemic hit, the world has been facing continuous health and economic issues. The art world, in particular, has been facing hardships that require art institutions to adjust their mode of operations. Since the year 2020, the world’s effort to contain the spread of COVID forced art galleries and museums around the world to close their doors and look for new forms of operation and exhibition. Such adaptations include increasing online marketing platforms, organizing virtual panels, and even creating online art exhibitions. In particular, these virtual exhibitions use high-resolution images of artworks, and provide them with contextual introductions of the artists’ background and inspiration. Some galleries include artworks that are available for sale, and thus further providing financial benefits for the galleries and their artists. The most fascinating part of these virtual platforms is the galleries’ implementation of virtual reality and augmented reality tools to produce virtual tours and remote immersive experiences. In other words, they are virtual exhibitions that mimic the audience’s experience when they are physically in an art gallery.

Virtual reality, also known as augmented reality (AR), usually displays an original or scanned work of art in a digital setting, thus creating a “total immersion” experience for the audience. As amazing and creative as it is for the audience, legal issues can arise for the gallery. For example, AR can invite “guerilla hacking” of a virtual exhibit. Hackers can copy and post unsanctioned works on the digital digital platform, and thus infringe upon the copyright of the original artists and take away the gallery’s potential revenue. Furthermore, the gallery also faces potential lawsuits from their artists alleging that the unauthorized use of their works was approved by the gallery.

As museums and galleries started implementing these virtual methods, they also had to start considering potential copyright issues. When museums use virtual reality or displaying art works online, they must keep in mind the intellectual property rights in the images and the text. Furthermore, they need to consider the rights of the artist, especially for a primary-market sale offer. For most artists, museums generally can clear the rights to use high-resolution images through the artist or her licensing agency. As for the photographer, if he or she is not employed by the artist or the museum, the museum should consider obtaining a broad license or require the photographer to execute a work-made-for-hire agreement with the customary in-the-alternative assignment language. Museums should also obtain the necessary rights from the author of the essays featured in the viewing room.

Museums and galleries may have available to them, the Fair Use defense against copyright infringement claims. For example, for secondary-market sales, such as resales of artworks, museums and galleries may not have a relationship with the artist or the artist’s estate. In such a case, the Fair Use Doctrine may allow the use of small, low-resolution images. The Copyright Act of 1976 provides that “the fair use of a copyrighted work is not an infringement of copyright.” To determine whether an allegedly infringing use is “fair use,” courts need to consider four factors: (1) the purpose and character of the use, including whether such use is of a commercial or for nonprofit educational purposes; (2) the nature of the copyrighted work; (3) the amount and substantiality of the portion used; and (4) the effect of the use upon the potential market for or value of the copyrighted work. Whether or not the doctrine allows the display of large-scale, high resolution images without permission is less clear. There is also no specific definition of large versus small scale, and high versus low resolution. Courts usually analyze each situation according to a totality of circumstances.

Lastly, galleries should be aware of whether or not the displayed artwork incorporates third-party content. If so, the owner of that content can potentially have a claim against the display. Possible solutions to mitigate this risk include obtaining an opinion from attorneys regarding potential Fair Use defense, working with the artist in advance of an exhibition to reach an agreement about the use,  and potentially having liability or omissions insurance in place. 

The online presence of museums and art galleries has grown due to COVID. Even now, after all venues have nearly reopened to the public, many virtual options still remain available. Although there are many uncertainties in potential copyright cases, museums and galleries that are using or considering virtual arts should conduct more thorough legal research, seek legal advice from counsel, and implement prevention mechanisms to mitigate risks.

Administrative Agencies & Their Role in Technological Regulation

/ Washington Journal of Law, Technology & Arts / Leave a comment
Photo by Andrea Piacquadio on Pexels.com

By: Chi Kim

On January 7, 2023, Kevin McCarthy became Speaker of the House after his colleagues from the House of Representatives held fifteen separate voting sessions. The House demonstrated an equally impressive and depressing feat given the inability of our current elected officials to achieve results for even seemingly mundane decisions. While many liberal observers may have rejoiced at the chaos, the fifteen votes is emblematic of an overall trend of inefficiency within the legislative branch and political processes, especially when tackling more fluid concepts and problems within the technology sector. Creating regulations requires large amounts of information, lobbying, and time to convince policymakers with inflexible positions and procedures around fluid and emerging technologies of the merits of the proposed regulations. In addition to the typical policy lag, the timeline for proposed technological regulations are further exacerbated by the following intrinsic and extrinsic factors. 

Intrinsically, Congress is not equipped to handle technological regulation by design. Although our most recent Congress is younger than its predecessor by one year, this small change alone is a historical anomaly. The 118th Congress is the third oldest since 1789 and generally has been climbing since the early 1980s.The average ages in the Senate and House are 63.9 and 57.5, respectively. While this could be the result of modern medical advancements, the increasing age of our elected officials bodes negatively for the hope that our policymakers will understand the technology that they are regulating. Remember, for instance, the famous Facebook hearings? Even the generally unpopular Mark Zuckerberg looked relatable when forced into the position of explaining a new technology to an older person. Beyond the general lack of subject matter expertise, congressional officials cannot invest the requisite time to learn about these issues while also tackling persistent issues within voting rights legislation, labor and supply chain constraints from international pressures, and a looming recession creeping closer layoff by layoff. 

Extrinsically, big tech still has a massive voice within our congressional chambers. During the 2020 election cycle fifteen major tech companies, including Amazon, Facebook, Google, Microsoft, Oracle, and others, spent $96.3 million to influence forthcoming bills like the National Defense Authorization Act, Fairness for High Skilled Immigrants Act, and the CHIPS for America Act. While Congress receives input from stakeholders, there is often a cost to frame their political positions. 

Despite our political gridlock, the American government is not completely unarmed against big tech. In political law, hydraulics is the concept that political energy is never destroyed but rather manifests into new forms, finding new gaps and openings within the regulatory or political landscape, much like water does on earth. In the context of the technological landscape, the responsibility of passing regulations has flowed to administrative bodies. The Federal Trade Commission (FTC), for example, influences technology policy in a number of different ways. The FTC recently filed a lawsuit against data broker Kochava Inc. for selling geolocation data from millions of mobile devices. If the FTC is successful, such a ruling would likely affect the overall data broker industry. Notably, the FTC leadership impacts the policy direction advanced by the agency. For FTC Commissioner, President Biden appointed Alvaro Bedoya, who previously served as the founding director of the Center on Privacy and Technology at Georgetown Law Center where he worked at the intersection of privacy and civil rights. Additionally, as of the writing of this article, the FTC is accepting public comments for a proposed rule to ban non-compete clauses. This rule is intended to increase worker earnings and create more competition among big tech. While administrative agencies do have their own procedural “policy lags,” the FTC can still actively tackle issues while receiving input from internal and external industry experts without being directly tainted by lobbying efforts. 

Law and technology are often portrayed as incompatible ideas — rising technology  meeting archaic regulations. However, policymakers need to realize that law and technology are not so different — both policymaking and technology development require troubleshooting and reiterations over time. However, unlike the software engineers in the companies that they regulate, policymakers do not have endless opportunities to sandbox their regulations before fully staking their political careers and capital. The responsibility of making such regulations has often flowed to administrative agencies that can take measured steps on the daunting task of regulating big tech companies. However, Congress should build on administrative agency efforts by passing bills based on the failures or successes of the agency actions. Doing so could result in more relevant and long-lasting technology regulations.