How to Get to Grips with AI Laws and Regulations

In recent years, there have been staggering technological advances and for most of us, it’s been a challenge trying to keep pace with the rapid changes. One of the more radical departures from the past has been the introduction and adoption of artificial intelligence or AI. This has transformed how many people work, it’s improved efficiency, empowered creatives and caused large numbers of people to lose their jobs. As you can imagine these changes have raised certain legal, moral and societal concerns about how AI is regulated and used. In this article, we will examine the rules surrounding AI in the US and the EU and how they could affect your life as we move forward. 

Do We Need AI Laws?

The short answer is yes. There is considerable scope for abuse and criminality. AI has been deployed by scammers and other disreputable people. We have already seen deep fakes of celebrities used in pornography or to push certain memes in the culture war. Text generation tools have been used to flood the online space with generic content that simply doesn’t feel right. Those that rely on AI for research have learned that it’s all too easy to make a mistake and ruin your reputation. The ever present threat of fake news has been exacerbated with deep fakes and it can be hard to trust any news report. Freedom of speech is vitally important, but there needs to be a way to regulate deep fakes to dissuade people from speaking lies and disinformation. For example: China has already put in place a mandate that any deepfake content must be accompanied with a disclaimer. This can be a good model to follow as we look at ways to control AI use that can be misleading. 

The Writers’ Strike

The recent strike by the Writers Guild of America (WGA) and the Screen Actors Guild - American Federation of Television and Radio Artists (SAG-AFTRA) brought some of these issues into focus. The strike was for pay increases that were unreasonable given the downturn in streaming and movie watching among the general public. Although SAG-AFTRA won higher pay, there are now more writers out of work due to a lack of demand. But, the new contract negotiations did include an AI clause to prevent the use of AI-generated text. The unions argued that AI-replication of content, including actors likenesses, voices and more would put them out of work. It is true that using deep fakes could, in theory, save a media company millions and increase profits. Paying an actor once to copy them and then using that AI copy in perpetuity without consent is obviously a step too far. Any future legislation concerning AI will almost certainly include clauses about consent. There will also need to be a risk assessment to ensure that a text generator isn’t stealing content that can be easily identified. This could in theory be an IP infringement that may lead to an expensive lawsuit in the future. 

5 Legislation Development Considerations

Let’s take a look at five important considerations that must be followed to create robust AI legislation:

• Algorithmic Bias: There need to be measures in place to prevent algorithmic bias to allow the AI to make unbiased decisions.

• Data Privacy: The regulations need to include how an AI can collect and use data to protect personal privacy.

• Ethical Principles: There should be fairness, transparency and accountability to encourage responsible AI use.

• Easy to Understand: The AI systems need to be easy to understand to ensure that users know how decisions are reached and implemented.

• International Cooperation: All governments should cooperate with designated international bodies to implement and enforce AI regulations.

How is AI Legislation Implemented in the EU and US?

To put all of this into perspective let’s take a look at how lawmakers are handling the implementation of AI legislation in the EU and the US specifically.

The European Union (EU)

There are plans for an “Artificial Intelligence Act'' to adopt a proactive approach to the implementation of AI legislation. Measures such as the General Data Protection Regulation (GDPR) are already in place to set guidelines for collecting, using and protecting personal data. The EU legislation will be aimed at the restriction of uncontrolled AI systems functions. This prospective AI Act will act in concert with the GDPR and it will give the EU control over how AI is developed, used and regulated. This AI Act is going to be guided by accountability, transparency and ethical principles to protect the interests of users. Hopefully the EU can leverage these principles to become a global leader in AI legislation and avoid the temptation to use this technology to grab more power on the continent. 

The United States (US)

The US has adopted a more decentralized AI regulation model that reflects how this nation governs itself. The policies and regulatory practices are focused on sectoral levels. At this time there is no regulation framework at the federal level, but there are sector-specific AI-regulated agencies that do address some AI evolution challenges. The National Highway Traffic Safety Administration (NHTSA) regulates AI directed autonomous vehicles. In a similar vein, the Federal Trade Commission (FTC) is more concerned with consumer protection issues and the use of AI in fair and transparent business practices. Certain stains have their own AI regulations, the CCPA has strict requirements on AI technologies and consumer data gathering. In summary, AI regulation in the US is not centralized, but it is equated with the extended sectoral part. 

In Conclusion

AI legislation and regulations will determine how this technology is used in the future to have a positive influence. The technology is advancing at a rapid pace and it’s likely that any legislation will need to be responsive to keep up with the possible ramifications. Coordination between the various regulatory bodies across national borders will be necessary. The Organization for Economic Co-operation and Development (OECD) at the UN may set and outline global regulations. Getting the most out of AI to solve the trillion dollar question that is not apparent at this time will only be possible with collaboration and shared responsibility for fair and just legislation.