Articles Posted in Intellectual Property

Machine learning algorithms can help create fake videos or pictures of someone else without their knowledge or consent. In fact, in 2017, University of Washington’s researchers created a video of Barak Obama who was seemingly discussing important issues. Now, there are software applications such as FakeApp that can help create deepfake pictures or videos for free. FakeApp was created by using Google’s open-source deep learning software program.

The advent of “fake news” has created a new movement in the entertainment and news industries. It has allowed everyone to question the source and validity of journalistic works. So now, deepfake movements and creations are creating new legal predicaments. The relevant issues, include, but may not be limited to, invasion of privacy, false light, and defamation.

The creator or publisher of the deepfake picture or video can put together a seamless video by having access to a base video and several source images of the person’s face. The computer-generated face can look identical to the original person’s face which can create confusion. This confusion can result in monetary damages to the victim. For example, a deepfake video can show the victim saying or doing something wrong which could cause the victim’s loss of employment. Or, in another example, the victim, who is running for political office, may be shown to have said or done something that could impede the election process.

Deepfake rules and regulations have been developing in the recent times. The term “deepfake” comes from two separate words – deep learning and fake – which uses artificial intelligence technology to create fake pictures or videos. The creator can utilize special software programs to create the picture or video by face swapping. This has become a problem because it can violate the victim’s privacy rights and public image.

We can detect the false image by conducting a reverse-image search. So, in other words, if the fake image was made by using another image on the web, the original version should be found. The fake image may be also detected by close evaluation. So, for example, the person in the fake video may not blink or yield normal facial expressions. It may also be detected through magnification or physiological analysis.

The victim’s legal rights can be violated by the deepfake creator or publisher. In most cases, it raises an issue regarding privacy rights. In California, false light is a legal cause of action that can be used by the plaintiff against the defendant who improperly represented the plaintiff who was embarrassed or offended by those actions. The plaintiff may argue that any reasonable person in the same or similar circumstances would be embarrassed or offended. The plaintiff may bring a cause of action for defamation against the creator and argue that the false factual statement – i.e., picture or video – was not privileged and had a tendency to damage his or her reputation in the community. The plaintiff may also file a legal action for misappropriation or right of publicity of the picture or video was utilized to promote a promote or service. Now, if the plaintiff suffers from emotional distress (e.g., depression, anxiety, insomnia), then he or she may also bring a cause of action for intentional or negligent infliction of emotional stress. See https://www.justia.com/trials-litigation/docs/caci/1600/1600 for more information.

Virtual reality (VR) is an experience taking place within simulated and immersive environments that can be similar to or completely different from the real world. It can be used for educational and entertainment purposes.  Also, augmented reality or mixed reality are other types of this technology.

The current systems use some kind of device (e.g., headsets or multi-projected environments) to create images, sounds, or other sensations in order to take the user into the virtual environment. The users can submerge themselves into the virtual environment and interact with the virtual features as if they are real.  The effect is generated by special headsets that yield a head-mounted display with a tiny screen before the eyes, but can also be created through specially designed rooms with big screens. This technology includes visual and auditory feedback, and can also permit other types of senses or feedback through haptic technology.

What are the potential legal issues?

We are continuing our look at the impact of artificial intelligence on copyright laws. In general, copyright’s underlying theory—to protect property rights as defined by the labor a person puts into an object, and to incentivize economic progress by entitling people to the fruits of their labor—will come to bear in face of this new technology.  It will happen in ways equally as novel as the technology at issue.

Although, copyright protection usually requires that a work be original and created by a human being, however, the fruits of artificial intelligence are becoming increasingly original.  It might prove unfair in some circumstances not to afford protection to a highly-complex and refined work of literature, music, or journalism simply because its creator is not human. It could likewise disincentivize the use of such innovation.  On the other hand, the argument is made that the market will disincentivize a human being from going through the laborious task of copying something that was previously made by a computer more efficiently.  The law really comes to a finer point when the issue is one of computer-generated copying of computer-generated works.  This thing intuitively feels unethical, but it is tricky to get the legal outcome that feels right under the current copyright regime. Another argument to be made in favor of not affording protection to works not created by human beings is that employing artificial intelligence to tackle time-consuming endeavors could really help the economy.

Another legal option besides denying copyright protection is to simply attribute authorship to the creator of the program that made the work being copied.  This manner of addressing the issue is practiced in places like Hong Kong, India, Ireland, New Zealand, and United Kingdom. The United Kingdom has a slightly different definition of authorship from that of places whose jurisprudence indicate that a human being is required: “In the case of literary, dramatic, musical or artistic work which is computer-generated, the author shall be taken to be the person by whom the arrangement necessary for the creation of the work are undertaken.”  This seems to allow for computer programmers to be characterized as authors, where in other countries, they may not be.

This month, our plan is to shift our focus on copyright laws, which at this time, is at a juncture of its own. Artificial intelligence is quickly sweeping across many areas of life.  Google, for example, has just started funding a program that will write local news articles. Perhaps more stirringly, depending on who you are, a group of Dutch museums and researchers unveiled The Next Rembrandt, a portrait created by a computer that had analyzed thousands of the artist’s works.  Google has its hand in the realm of art as well and is working on “Deep Mind” which is a subsidiary-developed software that can generate music by listening to recordings.

For many years computers have been able to generate crude works of art, as though they were tools like many others.  But today, artificial intelligence has reached new technological heights.  It is capable of learning on its own.  Autonomous systems may program themselves based on information they perceive, without the need of being programmed by human beings.  They can evolve and make independent decisions.  With regards to art, machines have not quite achieved the ability to “learn” something so subjective, but the neural networks softwares can generate highly-refined works. You may research Recurrent Neural Networks for more information.

One of the traditional requirements for copyright protection is that a work be original. One of the criteria used to determine whether a work is original is whether it was created by a human being. In fact, Europe and the United States of America have their rationale for this rule. In Europe, where copyright law is based more on a moral theory, centering on the rights of authors and the work, and even parts of their person they put into a piece.  In the United States of America, copyright law is based more on a utilitarian theory – i.e., society and government want to incentivize people to put in the labor and create valuable things with the knowledge that they will be rewarded with the fruits of their efforts, and others won’t be able to steal or unjustly piggyback off the work.