AI and IP Protection
Artificial Intelligence (AI), a concept that started being developed post-second world war, and is based on tranforming computing devices into smart machines, such that they become intelligent enough to perform natural human tasks, such as mathematical calculations, data processing, etc. AI technologies rely on the processing and computing capabilities of computing devices to learn and process data patterns and solve problems at a much faster rate than humans.
Therefore, because of this AI capabilities, its aggressively being used to mimic human intelligence, hence, the name artificial intelligence. Its artificial because they are human creation, based on computer logics and algorithms developed by humans. The sophistication and capabilities of a particular AI depends on the sophistication and capabilities of its developers. Based on AI’s capabilities to process large volume of data, it’s being used to process and analyse data, and “solve” problems. I personal believe AI is another tool in the hands of man to help him do his tasks faster. It’s still the task of humans. I will talk more about this in my subsequent write ups.
The issue am raising here is AI’s capabilities and implications for intellectual property. AI relies heavily on data provided to it; it does not have its own proprietarily generated data. It relies on the internet to process and analyse data. Think of Search engine 3.0 (or any level you want to call it). Based on the algorithm provided by its developers, its able to process available data and be able to deliver output (results) to the users. The danger here is the use of copyrighted or uncopyrighted materials available on the internet. Remember, AI is about learning and adapting data patterns in its analysis and output. Based on historical data patterns, AI is able to analyse provided data, and deliver the output-be it data projections, or research paper.
AI technologies are furiously disrupting all industries, most especially the entertainment industry. Now you can shoot a movie using AI characters, or even using AI to develop the scripts. Some are even using AI to generate popular actors’ voices and personae to make movies, and the audience, ignorant of this development, are loving it. AI is also used to illegally generate images/photos for people to use, a recent case between Getty and Stability, where Stability’s developed Stable Diffusion AI image generation system is used to access photos, part of whom are photos owned by Getty photos.
Another incidence of copyright violation is the one involving the mimicking of songs and videos of popular artists, as referenced in the action of Universal Music Group (UMG) pulling down AI song featuring fake Drake and Weeknd vocals from YouTube, Spotify, TikTok, Tidal and Apple. AI could be a tool for research, just like other search engine tools, but must be a tool that must be referenced, and not a complete copying tool, which is plagiarism. For example, how do you reference ChatGPT in a research work? Most especially, considering that ChatGPT does not own the material provided, these are curated content of some other people, writers and publishers who have taken the pain to develop these content, and are not being compensated for their works.
The quick defences normally raised by AI developers are fair use, transformative use, and de minimus, which are baseless in the light of public interest and violation of the rights of the copyright holders under the law. These are interesting times. Intellectual property is really under siege from AI developers, and its time for creators be aware of these developments, andensure the protection of their original works. They must work with their legal teams to be on the lookout for any infringement through the illegal usage of these works by AI generated tools.