The European Union (EU) has just passed the world’s first Artificial Intelligence law which seeks to provide regulations on its usage. While the law will apply only in the EU, experts who support the law hope for a “Brussels effect,” which is the phenomenon that laws and regulations passed in the EU are often passed in other jurisdictions. The Brussels effect was evident in Kenya with the passing of the Kenyan Data Protection Act shortly after the EU passed its data protection laws otherwise known as the GDPR.
As the use of artificial intelligence (AI) increases globally, human rights activists have increased lobbying arguing for regulation to curtail the negative impacts. Researchers and experts note that despite AI’s major benefits, it also poses very serious threats which could be dangerous to society unless curtailed by regulation.
The first negative impact is that AI may lead to job cuts in several industries. Many sectors are resorting to the use of AI due to its advantages over human resources. AI does not face physical limitations as human labour does. It does not get sick or want a day off and is accessible even after hours.
When a business uses AI it doesn’t face the challenges of complying with labour laws. AI is also much cheaper when compared to human labour where the business has to pay salaries and taxes associated with human resources.
Given these benefits, some businesses may opt to lay off staff and transition to AI. This will in the long term have severe repercussions on social welfare when there is a large number of unemployed people.
AI has created a new type of competition in the marketplace, where businesses compete for work with AI-driven solutions. In the graphic design and branding field for example, some businesses have opted to use AI driven solutions rather than procure the same service from suppliers. This type of competition may lead to what is categorised as “unfair competition.”
Secondly, AI may lead to the breach of data protection laws and intellectual property laws. This is not only a legal challenge but it also raises serious ethical considerations. AI may not be very accurate.
However, the most serious negative impact is if AI falls into the wrong hands, for example, using AI to profile victims during war. This will have very drastic consequences for humanity.
The discussions on the need to regulate AI have been going on for about 5 years and gained a lot of momentum with the advent of OpenAI. Lobbying for regulation increased after the launch of OpenAI’s ChatGPT in 2022, pitting human rights activists against tech companies. Some experts argued that AI is now a global issue and ought to be given the same risk status as global pandemics and nuclear wars.
The EU has led the world in enacting the first major AI law. The general policy thrust of the EU law is that it provides a human-centric and ethical approach to AI. The provisions seek to ensure human protection supersedes technological developments. As such, the law provides four types of regulation according to the level of risk posed by A.I. Level 1 is low risk and does not require regulatory intervention while Level 4 risk is unacceptable and banned.