UN Advisor Wants Human Rights-Centric AI Regulation

    Published on:

    As debate continues over the regulation of generative AI in the wake of this year's technology boom, a UN adviser is calling for governments to put human rights at the center of the parameters they adopt.

    Markus Klebsch, an expert in risk management and AI, neurotechnology and robotics and advisor to the United Nations, will focus specifically on safeguards to protect children while promoting sustainability and reducing the global digital divide. I think it is necessary to put

    Although there is a consensus on AI’s transformative capabilities; Proposed future methods This is because the government believes that machines should not be able to defeat humans. This relates to products that incorporate AI and other digital technologies.

    Also read: AI uses human brain cells to realize speech recognition

    AI-enabled products

    The proposal broadly covers AI-enabled products and other embedded digital technologies. These include internet-enabled toys, consumer drones, industrial robots, advanced medical diagnostic tools, smartphones, cameras, and smart speakers.

    This comes as AI gadgets have been criticized for potentially violating privacy, with privacy activists unearthing several children's toys and gadgets last month and secretly eavesdropping on children's conversations. Was.

    Other products The proposal also includes autonomous weapons, self-driving cars, and unembedded, standalone AI.

    “Smart devices such as home games, mobile phones, robots and drones are increasingly incorporating AI systems and could therefore be affected by these new regulatory proposals,” said Professor Emeritus at Stirling University of Management. Krebs said. school.

    “When governments develop regulatory frameworks, it is essential that they address our fundamental rights as citizens, from privacy, protection of individuals and data, to potential harm to the environment.”

    Concerns have been raised about personal data privacy violations by major companies such as Meta, which resulted in hefty fines in 2019. reaching $5 billion, It was warned by the Federal Trade Commission for violating user privacy.

    The company was also accused of violating users' privacy by failing to protect children's privacy with its Facebook Messenger Kids app.

    tough call

    Mr. Klebsch, who is also the founding director of the Human-AI Institute, has led the efforts of the United Nations Economic Commission for Europe Working Party 6 (UNECE WP.6). global guidance For member states from 2020 onwards.

    However, he acknowledged that developing global AI regulations for products that incorporate AI will not be easy and will be a challenge.

    “There are challenges in regulating products that incorporate AI. For example, products and technologies are continually evolving, raising questions about whether these products will remain compliant after firmware updates.” said Krebs.

    “However, I believe this proposal would allow member governments to develop their own horizontal digital regulatory frameworks for products that incorporate AI, reducing the risk to all stakeholders: policy makers, industry and the public. We hope to provide a solid foundation for not only mitigation, but greater certainty.”

    Transparency regulations

    Heidi Lund, Chair of UNECE WP.6, acknowledged that regulatory discussions targeting digital innovation and AI are ongoing around the world. These may vary by individual country and market, but must accommodate “a number of legitimate regulatory concerns related to digital products, including product safety, cybersecurity, resiliency, and privacy.”

    Regulation should help increase transparency and foster innovation.

    “Appropriate regulation should not be seen as stifling AI development,” Kebusch said.

    “In fact, the opposite is true: it will foster sustainable innovation aimed at ensuring secure infrastructure and reducing trade barriers for products that incorporate AI.”


    Leave a Reply

    Please enter your comment!
    Please enter your name here