As a specialist technology law team we have seen artificial intelligence (AI) and machine learning (ML) coming to the fore over the past few years. Here are some of the key trends and concerns that we’re seeing.
Not so long ago artificial intelligence was almost a marketing buzzword, rolled out by businesses across a wide range of sectors to enhance their offering. In reality, when you dug beneath the surface, many of these businesses were simply using sophisticated spreadsheets rather than any true machine learning. Increasingly though, this is not the case, with artificial intelligence finding its way into more and more everyday uses.
How are businesses using artificial intelligence?
Many businesses may find that they are using artificial intelligence technology without even realising it. Consumers engage with AI interfaces almost every day online and many times may not realise that they are doing so.
As with any emerging technology, there is a need to balance what can be done with the technology with what should be done with the technology both ethically and legally.
The law and ethics of artificial intelligence
It is a common trend among emerging technologies that technologists will push the boundaries of what society will find acceptable. Often it takes the lawmakers to put in place checks and balances to ensure that the technology, whatever that emerging technology may be, does not have a negative impact on society.
Almost inevitably, these legal boundaries are set in response to ethical concerns amongst the population.
In our experience, we are seeing a surge in both the prevalence of AI technology but also public awareness around what that AI technology is capable of. We have regulation on the horizon, driven by both ethical concerns and issues raised by the public.
Increasingly, as with good data governance, we anticipate consumers making decisions about their purchasing habits and interaction with companies based on how they interact with technology. This goes beyond whether or not they are using technology for a perceived good or just for commercial gain and focuses more on how the technology impacts on those consumers.
The current issues with artificial intelligence
Some of the keys issues that we are seeing at the moment and which we anticipate growing in the future include:
Automated decision making
Many businesses will find that they are making automated decisions about their interactions with consumers and businesses without even realising that they are doing so.
There are strict rules already in place around automated decision making when it impacts on data of individuals. Businesses should check whether or not what they are doing constitutes automated decision making and if it does will need make sure that they are complying with the relevant data protection legislation.
Access to decision making processes
Imagine that you’ve been the subject of an automated system, made by an organisation. Let’s say, for example, this is about whether or not an insurer will insure you. Imagine that you are negatively impacted by the decision and want to understand why that decision was reached. You very quickly will find that there is a tension between your rights to information about you and how you are impacted by the decision, and the interests of the technology provider who will be reluctant to reveal the inner workings of their technology.
These issues can be concerning enough when trying to find information about whether you are insurable (or your credit score, for example) but imagine how these issues could become even more serious as automated decision-making becomes more prevalent in day-to-day life.
This is not to say that automated decision-making doesn’t have its place, but organisations who embrace this technology need to be prepared for a growing trend towards transparency about how their technology works.
Balancing transparency with intellectual property protection
If businesses are to be more transparent about how they make decisions (and society is likely to demand it soon) then this creates a tension with how they protect the technology behind the decision-making processes. Businesses will need to become more imaginative about how they protect the intellectual property in their AI and ML activities.
Most readers will have seen common liability issues, which may arise as a result of artificial intelligence making decisions which give rise to a claim. The classic example would be a self-driving car that causes an injury. Whilst there is room for some debate here, the liability issues are generally much simpler than they may at first seem.
If a AI is considered as a tool helping a person complete a job just like any other tool then the liability for these sorts of scenarios would naturally sit with the person who has either developed that tool or provided training on it.
Traditional questions about negligence and responsibility soon come to the fore in these scenarios and so this may not prove to be quite the battleground that people expect.
Liability for IP infringements
Whilst the issues detailed above may or may not manifest themselves in a way that many commentators seem to think, there are almost certainly going to be issues around decisions made by AI programs which create breaches of intellectual property or which create intellectual property themselves.
Imagine for example, an AI program which develops text. If that text infringes somebody else’s intellectual property who will take the blame for this? Similarly, all intellectual property legislation was drafted on the assumption that intellectual property would be created by humans. When an AI creates new content which, had it been created by a human would be protectable by IP, there is some doubt as to who will own that intellectual property going forward.
Businesses will need to ensure that they are putting in place effective terms and conditions to ensure that whoever engages with their artificial intelligence programmes will understand the implications from an intellectual property point of view.
Last thoughts on artificial intelligence in 2021
From a legal point of view, the emergence of real world AI and ML technologies is exciting.
Much of the commentary today is focused on the wrong areas. The real commercial issues will be those which businesses find give consumers cause for concern and hence impact on consumers purchasing decisions. At the moment, those issues are likely to be around how a consumer’s data is used, the fairness of processing and the ability of consumers to challenge any decisions.
In future, we anticipate seeing an increase in issues arising as a result of both liability or damage caused by artificial intelligence and machine learning algorithms but also from content created by these technologies.
For more information about how we can help you to protect and commercialise your AI and ML technologies please contact us our specialist IP & IT team.