On the 1st August 2024, the EU AI ACT took effect. You’ll be forgiven for not noticing.
The AI Act is the first-ever legal and regulatory framework for artificial intelligence. The Act categorises AI systems into 4 different risk levels based on their potential impact on individuals and society.
Whether you believe it’s useful, over-hyped or a systemic risk to humanity, AI is being deployed everywhere. The Artificial Intelligence market in the UK is estimated to be worth $21bn, and estimated to grow to over $1tn by 2035, but until now there has been no regulatory framework in place anywhere.
While the UK has no AI regulation in place they hosted the AI Safety Summit in November 2023 during which the UK amongst 28 nations (plus the EU) signed the Bletchley Declaration. The declaration affirms that “AI should be designed, developed, deployed, and used, in a manner that is safe, in such a way as to be human-centric, trustworthy and responsible.
The recent change of UK government is unlikely to change the direction of travel towards regulation. The UK is however forecast to diverge from the EU AI ACT by expanding the 4 risk categories with 5 additional core principles that can be interpreted by regulators and applied variably, depending on industry sector, rather than just imposing inflexible technology-wide rules.
The additional 5 UK principles are:
Safety, security & robustness
Appropriate transparency and explainability
Fairness
Accountability and governance
Contestability and redress
On the surface, this expansion looks like a sensible move, as it more accurately assesses the AI’s use case, rather than limiting the law to what the AI function is.
A simple example of this would be the use of AI-driven delivery robots utilised inside a warehouse compared to AI-driven delivery robots to domestic homes. Clearly, there are health & safety risks & complexities that can occur in both settings, but these are exponentially greater once you are outside of a building. The principles above would allow regulators greater flexibility in the law to separately categorise the risk and governance of each use case.
In short, this benefits UK businesses in a few important ways:
Fit for purpose. UK businesses will be able to innovate AI solutions that benefit from greater legislative flexibility, maintaining safety & security without stifling their value and speed to market.
Protect your IP. The UK Regulation will also help establish a balance between AI innovation and copyright/IP law.
Better governance. Through the ICO, AI and Data Protection continues to be updated to clarify requirements for fairness in the use and deliverance of AI.
Reduce the risk of overregulation. The current EU AI Act places burdensome requirements (particularly for high-risk AI systems) on innovators & business. This could hinder and even deter UK R&D in AI.
Reduced costs. Many fear that strict compliance with existing AI regulations could drive up costs for businesses, especially smaller ones.
A fairer playing field. The EU’s strict regulations could very well place businesses at a competitive disadvantage compared to other global regions where regulations are less prohibitive.
What do we think?
At DOQEX, we have introduced and have been using LLM’s and AI to augment the efficiency, accuracy and utility of our security products. As such, we welcome the UK’s upcoming legislative approach and see it as an important step in the right direction in helping specialist technology businesses (such as ourselves) build greater customer trust, and compete against ‘big tech’ in this rapidly evolving technology space.
AI Regulation in the UK. How will it help you?
On the 1st August 2024, the EU AI ACT took effect. You’ll be forgiven for not noticing.
The AI Act is the first-ever legal and regulatory framework for artificial intelligence. The Act categorises AI systems into 4 different risk levels based on their potential impact on individuals and society.
Whether you believe it’s useful, over-hyped or a systemic risk to humanity, AI is being deployed everywhere. The Artificial Intelligence market in the UK is estimated to be worth $21bn, and estimated to grow to over $1tn by 2035, but until now there has been no regulatory framework in place anywhere.
While the UK has no AI regulation in place they hosted the AI Safety Summit in November 2023 during which the UK amongst 28 nations (plus the EU) signed the Bletchley Declaration. The declaration affirms that “AI should be designed, developed, deployed, and used, in a manner that is safe, in such a way as to be human-centric, trustworthy and responsible.
The recent change of UK government is unlikely to change the direction of travel towards regulation. The UK is however forecast to diverge from the EU AI ACT by expanding the 4 risk categories with 5 additional core principles that can be interpreted by regulators and applied variably, depending on industry sector, rather than just imposing inflexible technology-wide rules.
The additional 5 UK principles are:
On the surface, this expansion looks like a sensible move, as it more accurately assesses the AI’s use case, rather than limiting the law to what the AI function is.
A simple example of this would be the use of AI-driven delivery robots utilised inside a warehouse compared to AI-driven delivery robots to domestic homes. Clearly, there are health & safety risks & complexities that can occur in both settings, but these are exponentially greater once you are outside of a building. The principles above would allow regulators greater flexibility in the law to separately categorise the risk and governance of each use case.
In short, this benefits UK businesses in a few important ways:
What do we think?
At DOQEX, we have introduced and have been using LLM’s and AI to augment the efficiency, accuracy and utility of our security products. As such, we welcome the UK’s upcoming legislative approach and see it as an important step in the right direction in helping specialist technology businesses (such as ourselves) build greater customer trust, and compete against ‘big tech’ in this rapidly evolving technology space.
< Back to news & blogs
Want to know more?
Call Us
01904 500255
Message Us