We love new technology, but tools powered by artificial intelligence (AI) seem to create more questions than answer problems. While AI isn’t completely new, the launch of ChatGPT in November 2022 has reignited public interest in the tool which is currently undergoing a rapid improvement in quality. 

In this article, Squideo is going to break down why we should be cautious about AI-generated videos and how the law around AI is swiftly shifting.

This Just In

Since we first started looking at artificial intelligence (AI) in June, a lot has changed! At the time of writing in August 2023, here is an overview of the leading artificial intelligence headlines:

• An American court has ruled that AI-generated art cannot be copyrighted, stating “human authorship is an essential part of a valid copyright claim.” A related case is currently being processed by the UK Supreme Court.
• Artists have accused Adobe Stock of AI ‘copyright infringement’ as the AI art appears to mimic the style of the artist, at least partially.
• AI could revolutionise the mining industry, threaten the jobs of programmers, or possibly create the best racing game ever.
• A study by the United Nations (UN) has found that women face a disproportionate risk of losing their jobs to AI. This is because of their over-representation in clerical jobs, especially in the most economically developed countries.
• Governments around the world are racing to understand this new technology and regulate AI tools. 

These headlines suggest there is a need for caution when using AI-generators, but why is this new tool causing so many legal headaches?

The Risks of Artificial Intelligence

In the Compliance & Risk journal published by Osborne Clarke this month, it broke down the main legal risks businesses face when they use AI. This could be to produce written and visual content, and definitely applies if you’re using AI to generate a video.

01. Bias and Discrimination

A piece of work is only as good as its creator, and the same applies to AI. AI-generators are created by programmers who input data to train the AI. Over the past 20 years, we’ve seen plenty of times programmers have missed the mark. For example, in 2014 Amazon’s recruitment software was found to discriminate against female candidates applying for technical roles.

Programmers (or coders) can have unconscious bias or discriminatory attitudes. AI can also misinterpret data. Regulation of AI can help to identify this bias, yet that also comes with its own challenges. 

02. Government Involvement and AI Audits

Do we need government policies for how we use artificial intelligence? Is it up to the individual business to audit their AI, to understand potential risks and create risk mitigation strategies? Is it the responsibility of the AI creator? These are questions the world is grappling with, and the current laws surrounding AI could change significantly within the next 5 to 10 years.

Different countries are taking different approaches, much like they did during the evolution of data protection laws. For international companies, this means there could be many standards to take into account. Think of how the European Union’s introduction of GDPR changed the way websites around the world secure consent for data processing. Where AI saves time, it could also create more work for your company.

03. Intellectual Property and Copyright Infringement

Whether you’re training AI or using AI-generated content, there is a legal risk in using third-party intellectual property rights without securing permission or a licence. AI creators are facing private litigation, from individuals and businesses, to protect their intellectual property from unauthorised use. 

This month, it’s individual artists against Adobe Stock. In July it was Getty images against Stability AI. While text and data mining without permission is allowable in the UK, it does not allow data scraping for commercial purposes. How individuals and businesses will progress in these lawsuits is still to be decided, as a lot of cases remain in court.

As for work created by an AI-generator, it may not be copyrightable at all. In our headlines, there was an American court which ruled AI-generated art cannot be copyrighted since human authorship is essential for a copyright claim. AI can be used as a tool, but people must contribute to the creative process. For example, Squideo’s video producers could use an AI tool to alter images and the client would still have copyright over the video. Whereas if a client used an AI-generator to create an entire video, there would be no claim for legal ownership of the product.

04. Data Protection and Confidentiality

AI needs to be trained and, as discussed above, programmers often accomplish this with data scraping with information often sourced from the internet. What companies need to be wary of, especially if they are working within the UK or EU, is that data protection laws are not violated. 

“While information about, or images of, real people might be lawfully used to train AI in some jurisdictions, the practice faces challenges in the EU or UK, where data protection rules tend to be stronger than elsewhere.” Osborne Clarke, August 2023.

An AI-generated video might use the image of a person who has not given their consent to have this photograph shared, opening up the possibility of litigation. 

Work With Us

Video production companies like Squideo create your animated video from scratch, ensuring complete one-of-a-kindness. Ready to create a unique video of your own? Watch the video below to get a better understanding of how Squideo can help promote your business, then get in touch with us to find out more!