<img fetchpriority="high" decoding="async" src="https://technicalterrence.com/wp-content/uploads/2024/05/All-about-the-AI-regulatory-landscape.png" alt="All about the ai regulatory landscape” width=”100%”/>
Image from Canva
ai is advancing at a rapid pace, and while the possibilities are daunting to say the least, so are the risks that come with it, such as bias, data privacy, security, etc. The ideal approach is to have ethical and responsible guidelines. integrated into ai by design. It must be systematically built to filter out risks and only convey technological benefits.
Quoting Sales force:
“Ethics by design is the intentional process of incorporating our guiding principles of ethical and humane use into design and development.”
But it's easier said than done. Even developers find it difficult to decipher the complexity of ai algorithms, especially the emerging capabilities.
“According to deep controls“The skill in an LLM is considered emergent if it was not explicitly trained or expected during model development, but appears as the model increases in size and complexity.”
Since developers need help understanding the internals of algorithms and the reason behind their behavior and predictions, expecting authorities to understand them and keep them regulated in a short period of time is an excessive task.
Furthermore, it is equally difficult for everyone to keep up with the latest developments, leaving aside timely understanding to create adequate guardrails.
The EU ai Law
This brings us to discuss the European Union (EU) ai Law, a landmark measure that covers a comprehensive set of rules to promote trustworthy ai.
<img decoding="async" alt="All about the ai regulatory landscape” width=”100%” src=”https://technicalterrence.com/wp-content/uploads/2024/05/1715630983_720_All-about-the-AI-regulatory-landscape.png”/><img decoding="async" src="https://technicalterrence.com/wp-content/uploads/2024/05/1715630983_720_All-about-the-AI-regulatory-landscape.png" alt="All about the ai regulatory landscape” width=”100%”/>
Image from Canva
The legal framework aims to “ensure a high level of protection of health, safety, fundamental rights, democracy and the rule of law and the environment from the harmful effects of ai systems, while supporting innovation and the functioning of the internal market is improved.”
The EU is known for leading the way in data protection by introducing the General Data Protection Regulation (GDPR) previously and now for the regulation of ai with the ai Act.
The timeline
In the interest of the argument of why it takes so long to enact regulations, let's take a look at the timeline of the ai Law, which was first proposed by the European Commission in April 21 and later adopted by the European Council. in December'22. The tripartite dialogue between three legislative bodies (the European Commission, the Council and the Parliament) concluded with the EU Law coming into force on 24 March and expected to enter into force in May 2024.
Concerns Who?
With respect to the organizations that fall under its scope, the Law applies not only to developers within the EU but also to global suppliers that make their ai-act-a-primer/” target=”_blank” rel=”noopener”>ai systems available to EU users.
Risk classification
While not all risks are equal, the Act includes a risk-based approach that classifies applications into four categories: unacceptable, high, limited and minimal, based on their impact on a person's health and safety or fundamental rights. .
Risk grading implies that regulations become stricter and require greater oversight as the risk of enforcement increases. Bans apps that carry unacceptable risks, such as social scoring and biometric surveillance.
Unacceptable risks and high-risk artificial intelligence systems will be applicable six months and thirty-six months after the regulation comes into force.
Transparency
To start with the basics, it is crucial to define what constitutes an ai system. Keeping it too loose puts a wide range of traditional software systems under control as well, affecting innovation, while keeping it too tight can allow errors to occur.
For example, general-purpose generative ai applications or underlying models must provide the necessary information, such as training data, to ensure compliance with the Act. Increasingly powerful models will require additional details such as model assessments, systemic risk assessment and mitigation, and incident reporting.
Amid ai-generated content and interactions, it is difficult for the end user to understand when they see an ai-generated response. Therefore, the user must be notified when the result is not generated by humans or contains artificial images, audio or video.
Regular or not?
Technologies like ai, specifically GenAI, transcend borders and can potentially transform the way businesses operate today. The timing of the ai Act is appropriate and aligns well with the dawn of the era of generative ai, which tends to exacerbate risks.
With collective brainpower and intelligence, achieving ai safety should be on every organization's agenda. While other nations are contemplating introducing new regulations regarding ai risks or modifying existing ones to align them to handle the new emerging challenges of advanced ai systems, the ai Act serves as the gold standard for governing the ai. Establishes the path for other nations to follow and collaborate to put ai to proper use.
The regulatory landscape faces the challenge of leading the technological race between countries and is often seen as an impediment to obtaining a dominant global position.
However, if there should be a race, it would be great to see one where we compete to make ai safer for everyone and draw on ethical gold standards to launch the world's most trustworthy ai.
Vidhi Chugh is an ai strategist and digital transformation leader working at the intersection of product, science, and engineering to build scalable machine learning systems. She is an award-winning innovation leader, author and international speaker. Her mission is to democratize machine learning and break down the jargon so everyone can be a part of this transformation.
