Sat, Jun 22, 2024 page12

• INTERVIEW: AI regulation should focus on addressing risks: IBM

  • By Lisa Wang / Staff reporter
  • • 
    • 
    • 

  As Taiwan is accelerating its drafting of regulations on artificial intelligence (AI) following the European Council’s passage of the EU AI Act, IBM suggested that local regulators focus on regulating the use of the transformative technology to address various risks, including the generation of falsehoods and issues of bias.

  Paul Burton, general manager of IBM Asia-Pacific, offered his advice on AI regulations in an exclusive interview with the Taipei Times last week. During his 36-hour trip to Taiwan, Burton met with IBM’s key clients in the financial services and banking sectors.

  “I think there are certain principles that I think are good. One of them is not to regulate the development of technology, but to regulate the use of technology,” Burton said. “The regulation should never hinder the development of technology itself.”

  

  IBM’s share price is pictured on a screen at the New York Stock Exchange on Jan. 24.

  Photo: Bloomberg

  The Financial Supervisory Commission (FSC) on Thursday released its official guidelines, offering instructions for financial institutions and banks to introduce AI tools and manage potential risks based on four categories of AI life cycles. About one-third of 377 financial institutions, mostly banks, have adopted AI tools for customer services, risk evaluations and other functions, the FSC data showed.

  By the end of October, the National Science and Technology Council would be ready with a draft AI basic law for various businesses, as rapidly developing generative AI technology poses risks of intellectual property infringement, limited transparency in how the systems function, issues of bias and fairness, or security concerns.

  “I do not want to recommend any for Taiwan. Taiwan should come up with its own laws,” Burton said when asked if he would suggest any references.

  With implementation of the EU AI Act imminent, many countries are moving ahead with robust regulations and guidelines, but business leaders around the world are feeling pressure to prepare, Burton said.

  They are not exactly sure what they are preparing for, he said.

  Based on IBM’s “2024 CEO Study,” few than 60 percent of global executives think their organizations are prepared for AI regulation and 69 percent expect a regulatory fine due to generative AI adoption, Burton said.

  In the face of this uncertainty and risk, CEOs are pumping the brakes, he said.

  More than half, or 56 percent, are delaying major investments until they have clarity on AI standards and regulations, he said.

  Overall, 72 percent of executives say their organizations will forgo generative AI benefits due to ethical concerns, he said.

  As the financial service sector is IBM’s cash cow, the company launched the Watsonx platform, opening up exciting prospects in financial advisory and data analysis, Burton said.

  Configuring generative AI models correctly can simplify complex financial concepts and make them easier for customers to understand in deterministic financial environments, he said.

  Financial institutions need to use generative AI thoughtfully to find the right balance between innovation and ethical practices, he added.

  Banking is among the industries that could see the biggest impact as a percentage of their revenues from generative AI, especially in terms of customer operations, marketing and sales, banking software, and risk and legal compliance, Burton said.

  Across the banking industry, for example, the technology could deliver value equal to an additional US$200 billion to US$340 billion annually if just currently identified use cases are fully implemented, he said, citing the estimates by the World Economic Forum last year.

  As IBM is at the forefront of technologies like AI, hybrid cloud and quantum computing, the company built a powerful AI governance into its Watsonx platform and developed quantum-safe cryptography to secure sensitive data, he said.

  It is why IBM advocates for smart AI regulation, including holding those who develop and deploy AI accountable for fraudulent, discriminatory and harmful activity, he said.

  It is also why IBM and Meta announced the formation of the AI Alliance, a group of more than 70 organizations dedicated to advancing open, safe and responsible AI, he said.
• Most Popular

  • 1

    Roadside saliva tests to begin by end of the year

  • 2

    Registration for NT$10,000 to open on Nov. 5

  • 3

    MOFA rejects ‘Time’ ‘reckless leader’ label

  • 4

    Taiwan in Time: The breakfast that sparked Taiwan’s semiconductor revolution

  • 5

    NT$10,000 cash handout explainer