Home » ISO Articles » A Legal Approach to Artificial Intelligence
A Legal Approach to Artificial Intelligence
Overview
One of the most significant technological developments in recent years is the creation of artificial intelligence, or AI. This is becoming more and more pervasive in everyday life and work. ‘Virtual assistants’ such as ChatGPT are being used by millions to help in multiple situations, such as content creation, idea generation and problem solving. Many AI models are being used to generate images, with the user being able to input prompts which are then used by the AI to create visual content.
In many ways, AI is a great tool that can be used to make our lives easier. It can help businesses run smoothly and can be implemented in many areas to improve functionality. However, the increase in AI and the potential risks it brings along with it means that some form of legal regulation is highly likely to happen, and in some cases, it is in demand.
Present and Future Artificial Intelligence Regulation
As of writing, there is no current legislation specifically around artificial intelligence in UK law. This is expected to change in the future as it becomes increasingly necessary to regulate AI use, especially among businesses
The EU Artificial Intelligence Act, otherwise known as Regulation (EU) 2024/1689, aims to form a comprehensive framework for AI use across the EU. It introduced a classification system to measure the risk levels of artificial intelligence application, with the categories being minimal, limited, high and unacceptable levels of risk. For minimal risk applications, such as spam filters, minimal regulatory oversight is needed, although codes of conduct may be adopted. Limited risk applications include chatbots, in which transparency measures are needed, such as making it clear when users are interacting with AI.
For high-risk AI applications, the most amount of regulation is required. This covers AI systems being used in areas such as healthcare or critical infrastructure. Risk management requirements include maintaining detailed documentation and implementing risk management processes. Businesses also need to make sure their systems can be audited for compliance, with clear information on how AI decisions are made and managing data in a way that upholds data privacy. Non-compliance issues and incidents must be reported, and businesses should participate in regular monitoring exercises.
At the unacceptable risk level, the act outlines banned uses of AI. Any uses of AI considered as a threat to safety, fundamental rights or involving social scoring are banned and cannot be practiced.
Implications of AI Legislation
Keeping up to date with AI legislation has many benefits if your business utilises artificial intelligence. The clear and defined risk management approach along with the risk level classification system can be used to reduce risk around the use of AI. The legislation ensures that AI being used is reliable and trustworthy, which may be especially important for safety when using AI created by a third party. Finally, transparency around how AI is used by your business has a positive effect on customer trust, allowing your business’s AI usage to stand out.
The Future of AI Legislation?
Other than using the EU Artificial Intelligence Act as a guideline of what legislation we may expect to see in the UK, there could be many other ways in which artificial intelligence is regulated and monitored. As AI can be used for data handling, there is overlap with data protection legislation that could be outlined in future legislation to apply specifically to AI data processing and storage.
Another major concern for artificial intelligence is protecting intellectual property. AI systems must be trained using a database of existing works, causing some dispute when creators do not consent to having their works used to train AI. This could call for legislation around protecting intellectual property and requiring clear information about what artificial intelligence systems are being trained with.
Although this is speculative, it is important to keep up to date with the progression of artificial intelligence legislation to ensure that your business can experience the full benefits of AI systems with minimal risk.