What Will AI Regulation Look Like for Businesses?

Unlike sectors such as food, medicine, and automobiles, the realm of artificial intelligence (AI) lacks clear regulations or legal frameworks in the U.S. This absence has led companies creating machine learning (ML) models to rely on their own interpretations of ethical guidelines. However, this situation is on the verge of transformation.

With the European Union (EU) finalizing its AI Act and the swift evolution of generative AI, the regulatory landscape for artificial intelligence is shifting from broad recommendations to more concrete laws. The EU AI Act has spurred significant discussions among business leaders focused on preparation for more stringent AI regulations, concerns about proactively designing compliant AI systems, and speculation on when similar laws might emerge in the U.S. This article will provide insights into what AI regulations may entail for companies in the near future.

Impact of the EU AI Act on Your Business

Similar to the EU’s General Data Protection Regulation (GDPR) implemented in 2018, the EU AI Act is poised to set a global benchmark for AI governance. The European Parliament is anticipated to vote on this draft by the end of March 2023, with potential adoption by year’s end if the timeline is adhered to. There’s a strong likelihood that the ramifications of the AI Act will extend beyond EU borders, leading international organizations to comply with this legislation. Furthermore, U.S. companies and other independently-led firms will likely recognize the advantages of adhering to these new regulations.

Observations of similar legislative proposals—like Canada’s Artificial Intelligence & Data Act and New York City’s automated employment regulation—further support this trend.

AI System Risk Categories

The AI Act will classify AI systems into three risk categories, each with specific guidelines and consequences:

• Unacceptable Risk: AI systems in this category will be prohibited. This includes manipulative technologies that cause harm, real-time biometric identification systems for law enforcement in public areas, and all forms of social scoring.
• High Risk: This category encompasses AI tools such as job applicant screening systems, which will be subjected to particular legal requirements.
• Limited and Minimal Risk: Many current AI applications, like chatbots and AI-driven inventory management, fall into this category and will mostly be unregulated. However, customer-facing applications in this category will require disclosure regarding their AI usage.

Looking Ahead: AI Regulation

Since the AI Act remains in draft form and its global implications are yet to be fully understood, it is difficult to predict the specific regulations organizations will face. However, future regulations are expected to vary based on industry, the type of model being developed, and its assigned risk category.

Regulation will likely involve third-party scrutiny, where models are rigorously tested against the populations they serve. Questions during these evaluations may include “Is the model operating within acceptable error margins?” and “Are you transparent about the model’s nature and use?” For organizations with high-risk AI systems, the AI Act has already delineated several obligations:

• Implementation of a risk-management system.
• Data governance and management practices.
• Technical documentation requirements.
• Record keeping and logging protocols.
• Transparency and information provision to users.
• Human oversight mechanisms.
• Standards for accuracy, robustness, and cybersecurity.
• Conformity assessments.
• Registration with relevant EU member-state governments.
• Post-market monitoring systems.

Regular reliability testing for AI models could also become a standard service within the industry, much like vehicle e-checks.

Preparing for AI Regulations

Many AI leaders are increasingly prioritizing trust and risk mitigation during the design and development of ML models. Embracing AI regulation as a new norm is crucial for forward-thinking organizations aiming to thrive in this changing landscape.

Explore a wide variety of countries and territories across the globe. Here is a comprehensive list:

Islands: Faroe Islands, Fiji, French Polynesia, French Southern Territories, Greenland, Guam, Isle of Man, Kiribati, Montserrat, Norfolk Island, Pitcairn, Samoa, Solomon Islands, Tokelau, and many more.

Countries: Gabon, Gambia, Georgia, Germany, Ghana, Greece, Guatemala, Guinea, Haiti, India, Indonesia, Iran, Iraq, Ireland, Israel, Italy, Jamaica, Japan, Jordan, Kazakhstan, Kenya, Kuwait, Latvia, Lebanon, Lesotho, Liberia, Libya, Lithuania, Luxembourg, Madagascar, Malawi, Malaysia, Maldives, Malta, Mexico, Monaco, Mozambique, Myanmar, Namibia, Nepal, Netherlands, New Zealand, Nicaragua, Nigeria, North Macedonia, Norway, Oman, Pakistan, Panama, Peru, Philippines, Poland, Portugal, Qatar, Romania, Russian Federation, Rwanda, Saudi Arabia, Senegal, Singapore, Slovakia, Slovenia, South Africa, Spain, Sri Lanka, Sudan, Suriname, Sweden, Switzerland, Syria, Tajikistan, Tanzania, Thailand, Tunisia, Ukraine, United Kingdom, United States, Uruguay, and Uzbekistan.

Other Regions: French Guiana, Guadeloupe, Holy See, Gibraltar, Mayotte, New Caledonia, Puerto Rico, Saint Barthélemy, Saint Kitts and Nevis, Saint Lucia, Saint Vincent and the Grenadines, Turks and Caicos Islands, US Minor Outlying Islands, and more.

Lastly, don’t overlook the territories like the Virgin Islands (British and U.S.), Western Sahara, and the areas of Wallis and Futuna.