The urgency to regulate powerful AI has increased in recent times, as governments scramble to address the risks it might create. The European Union, for instance, passed landmark AI regulations last month, amid concerns that the powerful tech grabbing everyone’s attention needs to be reined in. But strides made in one region may hardly be enough to tackle AI’s challenges.
Instead, global leaders should work together if they hope to arrive at the “right solution,” Pablo Hernández de Cos, the chair of watchdog Basel Committee on Banking Supervision and Bank of Spain governor, told the Financial Times in comments published Monday. If they don’t, AI could “change the course of history not necessarily for the good,” he warned.
“If we are not able to give a co-ordinated global response, the likelihood of getting the right solution to these challenges will be reduced,” de Cos said.
The governor spoke about financial institutions as an example of how global cooperation can help make systems more robust and safe.
“The fact that we are keeping this degree of coordination on financial stability is really remarkable, in particular because in other domains we are not seeing the same and perhaps we can to a certain extent leverage from this experience,” de Cos said in comments from the FT interview published on the BCBS website.
A similar approach as that of the world’s financial regulators should be applied to AI, de Cos suggested.
“We know that it is impossible to solve the challenges posed by topics such as AI or climate without a common international agreement,” the financial watchdog’s chair said, adding that financial stability was just one aspect where AI’s impact can trickle down and potentially result in bad consequences.
“Financial stability is only one dimension, there are many other potentially more important
consequences related to AI. Issues that if not properly managed could change the course
of history not necessarily for the good,” de Cos said.
The Spanish governor’s comments comes just ahead of the World Economic Forum’s kick-off in Davos, where AI’s promises and risks will be a key theme of discussion. Some of the industry’s most influential figures including Meta’s chief AI scientist Yann LeCun, Microsoft CEO Satya Nadella and Inflection AI cofounder Mustafa Suleyman are to speak at the event. The cutting-edge tech’s regulation, which forms an important piece of the polarizing debate about controlling AI, is expected to be among the topics of discussion.
AI frameworks in the making
Although some of the biggest tech companies, which are also among those driving the development of AI, have put their own guardrails in place, a groundswell of experts have called for more frameworks to ensure the tech isn’t used to society’s detriment. Some experts have sounded the alarm on the possibility of generative AI being misused in elections while others have suggested ways to regulate it by registering AI tools so governments can track them.
Last year, a group of tech luminaries came together to sign a letter urging a six-month pause of the development of AI tools, calling for a dramatic acceleration in “development of AI governance systems” and government-established bodies to deal with the disruption the tech causes. While the moratorium was never put in place, the letter fueled the debate on whether AI’s progress needs to be slowed and how involved governments need to be.
Tech companies that lie at the center of AI’s proliferation have supported the need for regulations while arguing for it to be “balanced” given the tech’s fast-evolving nature.
“We think policymakers, academics, civil society, and industry should all work together to minimize the potential risks of this new technology, but also to maximize the potential benefits,” Meta CEO Mark Zuckerberg said during a U.S. Senate committee hearing in September. Tesla CEO Elon Musk also said at the time that the U.S. needed a “referee” in the context of AI, while Google boss Sundar Pichai has also highlighted the need for “international cooperation” as the path forward.
Although the U.S. government has been pushing to make progress with developing the necessary frameworks to govern AI, the country has been lagging behind Europe in achieving this. But in a recent sign of global collaboration, America, Britain, Germany and a dozen other countries signed an agreement in November on AI safety that advocated for AI creations to be “secure by design.”
It remains to be seen how cooperation at the global scale might play out in the AI realm, but it’s safe to say the stakes are high and everyone is waiting with bated breath to see how it unfolds.