IP Protection, Copyright Issues With AI-Generated Work

Share This Post

We recently talked about how AI is taking over the marketing industry by storm in our previous story AI Will Disrupt the Content Marketing Space in 2022 & Beyond. Content marketing and Artificial intelligence are also used to create everything from software to paintings and music (Suggested Reading: AI in Marketing: Strategies, Examples, and Everything You Need to Know). 

However, it seems like businesses and marketers will need to keep a human staff to ensure that IP regulations protect their machine-generated inventions.

The US Copyright Office recently denied an AI-generated work of art (1). And the three-member panel determined that any artistic creation must have an element of human authorship.

“The connection between the human intellect and creative expression,” according to the board’s decision, is a crucial feature of copyright.

The Deal Breaker

Dr. Steven Thaler, a US scientist, intended to patent an algorithm he calls the “Creativity Machine,” which reprocesses photographs to produce hallucinating visions for a “simulated near-death experience (2).” Despite classing the algorithm as “work for hire,” the Copyright Office rejected it because the AI works with minimum human intervention.

Although copyright law does not specifically establish guidelines for non-humans, courts have viewed assertions that animals or divine entities can benefit from copyright protections with suspicion. According to a 1997 ruling, a book of (claimed) divine revelations (3), for example, could be protected if it contained (allegedly) human arrangement and curation. 

A court ruled in 2018 that a monkey could not file a copyright infringement lawsuit (4). The board claims that “the courts have consistently found that non-human expression is ineligible for copyright protection.”

Further investment in automated systems for art and the creative economy could be stifled as a result of this condition.

However, it doesn’t indicate that any art piece made via AI is not eligible. 

Dr. Thaler emphasized that humans were not involved in his project in any way because his aim was to demonstrate that machine-created works could be protected, not simply to prevent individuals from stealing the image. (He unsuccessfully attempted to show that AIs can also patent inventions in the United States (5)).

However, the board’s final decision appeared to have taken his explanation for granted.

It means that if anyone attempted to claim copyright for a similar work listing the product of their own creativity done by a computer, we might see a different result. At the same time, if Dr. Thaler files a lawsuit after this rejection, a court may come to a different conclusion about his work. 

Nonetheless, with machine-produced works, the Copyright Office emphasizes the relevance of human agency. The limits of that conclusion could be explored for years to come as AI becomes more prevalent in artists’ repertoires.

Read Also: 5G Technology: Advantages, Facts, Concerns, and Myths

The Difference Between Copyright and Patent

Despite the blow, Dr. Thaler acquired the first-ever patent in Australia for two inventions generated by his AI system DABUS (6).

Because the Australian Patent Act (Act) prohibits the artificial intelligence machine from being considered an inventor, the Australian patent office determined that such a system could not be an inventor. On the other hand, the judge found the artificial intelligence system to an inventor for the purposes of the Act (7).

The judge first reasoned that an inventor is an agent noun, meaning that an agent can be a person or object that invents. As a result, an artificial intelligence system can be described as an “inventor” if it is the agent that invents. 

Furthermore, nothing in the Act mandates the other conclusion. The judge went on to say that there is no particular provision in the Act that clearly prohibits an artificial intelligence system from being an inventor. 

In other words, unlike copyright law, which requires a human author or the presence of moral rights, there is no specific characteristic of patent law that would lead to an interpretation of the Act that excludes non-human inventors.

The judge then reasoned that it is consistent with the Act’s goal to interpret the term “inventor” in a way that encourages technical innovation and its publication and diffusion by rewarding it, regardless of whether the innovation was created by a human or not. 

Without the opportunity to get patent protection, the judge reasoned, owners of creative computers would opt to hide patentable ideas as trade secrets, keeping them hidden from the public eye.

Furthermore, the applicant is the owner, programmer, and operator of DABUS, the artificial intelligence system that produced the invention, and machines have been utilized to develop patentable results autonomously or semi-autonomously for some time. 

It means the applicant, as the owner and controller of DABUS, would own any DABUS innovations that came into his hands and would be entitled to have the patent issued to him as the owner of the invention. And the Act only stipulates that the applicant has the right to have a patent assigned to him if the application is granted.

As a result, the judge determined that an artificial intelligence system or device could be an inventor under the Act. However, a non-human inventor cannot file a patent application or receive a patent (89).

His patent applications in the United States and the European Union, meanwhile, are still pending.

As discussed, the outcome of those cases, and others like it, will have far-reaching repercussions not only for artificial intelligence inventor patent applications but also for other areas of patent law such as conception and obviousness.

In terms of copyrights, because AI is used to create music, news broadcasts, and games, such works will continue to be free to use and reuse.

Read Also: Biohacking: An Industry With Opportunities Worth Over $50 Billion

Who is Dr. Stephen Thaler?

Dr. Stephen Thaler is a Ph.D. physicist who has worked on stealth and nuclear technologies, high-energy lasers, and the laser-driven production of novel materials, among other things (10). His most significant scientific contribution is the finding that various types of random physical disturbances within the brain drive both cognition and consciousness. 

Many of Dr. Thaler’s dozens of artificial intelligence patents have been inspired by this work, which is largely focused on encouraging these chaotic brain models, which he refers to as “Creativity Machines®,” to develop and create much beyond their direct experience.

Through his examination of dying neural nets creating simple text, music, and pictures, his early investigations into brain function delved into the most disputed cognitive phenomenon, near-death experience (NDE). A brain simulation, dubbed “DABUS,” has developed itself using neural network techniques he pioneered over the last three decades, with more neurons than its human ancestor (11). 

This vast cerebral system continues to produce increasingly complex works of art and verbal wit. In the meantime, lower levels of disruption in these synthetic brains have resulted in various practical outcomes. 

The invention of new consumer products, synthesis of novel and technologically useful materials, control of constellations of communication satellites, improvisational battlefield robots, and highly intelligent Internet spiders searching the internet for signs of terrorist activity are just a few examples of these accomplishments.

Read Also: Google’s Time Crystals Demonstration: Layman Explanation

Closing Remarks

Overall, the hardware and software developers who use AI and want their work copyrighted will have to maintain human input to qualify. 

If you are interested to read more about the nature of intelligence, we found this interesting paper published in Nature, a peer-reviewed journal about deep reinforcement learning techniques introduced by Google Brain last year (12). It focuses on floorplanning, the process of arranging the placement of the different computer chips components.

With the reinforcement learning technique, the researchers managed to design the next generation of Tensor Processing Units, specialized artificial intelligence processors of Google (1314). 

Of course, the use of chip design is not new. But this reinforcement learning-power chip design indicates the need for abstract thinking, developing intuitions about the solutions to the right problems, and choosing the correct data for the validation to innovate AI software and hardware. And these are the skills that better AI chips can enhance but will probably never replace. 

At the end of the day, we believe it is about the manifestation of humans finding ways to use artificial intelligence as a prop to overcome their cognitive limits and enhance their capabilities (15). 

And it all points to a positive synergy between AI and human workers, as companies will require human laborers attached to a project alongside AI to secure its copyright protection.

+ posts

Rucha Joshi, currently managing a team of over 20 content writers at TimesNext is fueled by her passion for creative writing. She is eager to turn information into action. With her hunger for knowledge, she considers herself a forever student and a passionate leader.

spot_img

Related Posts

Niti Aayog floats Digital Banks

  Niti Aayog has proposed setting up full-stack digital banks...

Swiggy allows Work From home

Joining the list of startups Mesho, Spinny, Zerodha, and...

Can A.I. cure Cancer? 

When we hear the word "terminal disease," the one...

Metaverse in Medicine

Since its inception, Metaverse has been such a buzzword. From...

Monkey Pox – Another endemic?

The world was having a sigh of relief after...

KGF verse – All you need to know!

After a resounding success of KGF, chapter 1, the...