Industry executives and experts share their predictions for 2024. Read them in this 16th annual VMblog.com series exclusive.
AI Regulation and Innovation in the Year Ahead
By Natália Fritzen, AI Policy and Compliance Specialist at Sumsub
There’s a general consensus that without proper guardrails in place, AI-powered technologies will fail to deliver on their promise to generate societal gains. Since the release of ChatGPT in late 2022, AI has become a topic du jour. Specifically, the conversation centered around regulations: what should they be, how should they be implemented, and how to provide enough guardrails without stifling innovation? AI presents a challenge for governments and regulators, as something changing so quickly defies the traditional regulations made to stand the test of time.
While President Biden’s Executive Order (EO), the EU AI Act, and the UK government-sponsored AI Summit marked important milestones for the AI industry this year, we have not yet reached a ‘finish line,’ and step-by-step guidance is needed to properly regulate AI technologies while encouraging innovation.
Companies must truly understand and abide by the underlying requirements of upcoming AI regulations to operate successfully
Moving past the AI regulatory buzz in the United States and the European Union this past year, both regions will finally see real outcomes as bills regarding AI regulation are presented and voted on by the American Congress, and the EU AI Act is set to be approved come early 2024.
At the same time, as AI regulation will be a leading element of regulatory affairs in 2024, such a challenging landscape will impose a need for companies to understand the principles and requirements underlying AI rules, especially those applicable to the markets in which they operate. As such, AI compliance specialists are an integral new role for companies who will need to embrace AI safety as it is set to become an integral part of their activities or risk getting left behind.
The US will play catch up, mirroring the current efforts of other countries
Until now, the US treated guidance for the use of AI as precisely that, recommended guidance. Even the Biden Administration’s EO doesn’t mean enforced regulation is underway; rather, it is officially on the table. This laissez-faire approach doesn’t compare to what other parts of the world are already doing, and the US must model the participation of different stakeholders in the regulatory process and incentivize safe innovation to see true change.
Now tackling forward-looking, more daunting topics like frontier AI and AI existential risk, The UK House of Lords recently presented its AI (Regulation) Bill, which is a step towards implementing some of the provisions of the UK AI White Paper, a non-binding guidance material the UK government issued in March 2023. However, for any region, the first step needs to be addressing existing issues, such as AI bias or discrimination, before they can even think about solving the others.
Addressing AI safety will become a priority, demanding greater collaboration among stakeholders
In the coming year, it’s likely more AI models will contribute to the rise of deepfakes. As AI technology advances, collaboration among stakeholders remains essential for detecting and combating deepfakes in order to prevent the spread of misinformation. The AI community will realize the importance of engaging in ongoing discussions to help identify and reduce the use of deepfakes, as it is a never-ending and evolving process.
Additionally, Sumsub’s Identity Fraud Report found AI-powered fraud, including deepfakes, were a top fraud type in 2023, increasing 10x globally, which further emphasizes why multiple perspectives and shared information are crucial to protect consumers and businesses as more sophisticated forms of fraud emerge.
In 2024 and beyond, society needs to walk the fine line between future and existing issues in order to truly shape AI governance and stop viewing regulations as a roadblock hindering innovation, rather as the next step in AI innovation getting the care it warrants.
##
ABOUT THE AUTHOR
Natalia Fritzen is the AI Policy and Compliance Specialist at Sumsub, where she supports the company in navigating the world of AI regulation. Prior to her current role, Natalia served as a tech policy analyst at Nubank. She holds a Master’s degree from the Central European University, where she specialized in Ethics and wrote her final thesis on AI nudges and regulation.
