Artificial intelligence is well past early stages. It has evolved rapidly, is already shaping the future of almost every industry and its growth prospects are phenomenal. Take ChatGPT for instance. By February 2023, it was the fastest growing app in internet history, with 100 million users and 25 million daily visits. The global AI market, valued at $136.55 billion in 2022, is expected to reach a whopping $1.8 trillion by 2030.
We don’t really need statistics to gauge the power of AI. It’s evident from the EU’s readiness to regulate it. The EU has made a swift first move again. Not in innovation or technology, but in announcing the world’s first regulation on artificial intelligence – the EU AI Act. And, no, tech giants aren’t thrilled. Having poured billions of dollars into R&D, the world’s biggest companies have reason to worry about an Act that aims for a “safe, transparent, traceable, non-discriminatory, and environmentally friendly” AI ecosystem. That could translate to a bunch of highly restrictive regulations, which is reflected in the bearish sentiment for tech stocks in Acuity’s AssetIQ widget.
In May 2023, McAfee released a report titled Artificial Imposter to highlight a “new breed of scams” perpetrated by voice-cloning AI tools. AI undoubtedly poses threats and needs to be regulated. The question is whether it’s time for a highly restrictive set of rules that will stifle growth.
In early 2021, the European Commission had proposed a regulatory framework that classified AI systems as per the potential risk they posed to users. These classifications included unacceptable risk, high risk, and limited risk. Since the spotlight has moved from Gen Z to Gen AI, or Generative Artificial Intelligence, this came under a classification of its own.
AI systems identified to pose unacceptable risk are proposed to be banned. These include voice-activated toys that could lead to undesirable behaviours among children, social scoring applications and even biometric ID systems, like facial or voice recognition.
AI systems identified to pose high risk are proposed to be closely monitored. These include an array of segments, including medical devices, education, toys, radio equipment, marine equipment, motor vehicles, aviation and rail.
Several restrictions are proposed to prevent Gen AI from violating copyright rules. Also, before launching any AI system to users, companies will be forced to carry out a “fundamental right impact assessment.”
AI holds the promise of safer transport, better healthcare for more people, cleaner manufacturing, sustainable energy, and lower prices all around. The EU AI Act will pull the plug on many of these developments penetrating the bloc.
While the EU has again taken a risk-first approach, the UK is leaning toward a pro-innovation one. The US is still grappling with defining the impact area of the technology. In the highly articulate ways of Americans, Senate Majority Leader Charles Schumer recently said that they were still months away from looking into “specific stuff” to propose a framework.
The biggest fear among the tech giants is that the EU regulations may serve as a baseline for other countries. If other major and emerging markets follow suit, it may be decades before tech companies recover their huge investments in AI. This includes some of the world’s biggest companies, like Microsoft, Apple, Amazon, Google, IBM, Nvidia, Intel, Salesforce, and Meta Platforms.
The stringent rules and close scrutiny proposed by the EU AI Act could either prevent tech behemoths from offering their full range of services to member nations or spell millions of dollars in compliance costs.
Tech majors responded quickly, with 160 executives signing an open letter criticising the Act on June 30, 2023. Leaders in their domains, like Renault, Heineken, Airbus, and Siemens, were all part of the signing committee. The letter urges the EU to follow an approach that favours technological advancements.
Dragos Tudorache, one of the members of the drafting committee, said that the proposal is not yet fully understood and that it will be during upcoming discussions. Maybe there’s some hope there.
If the world’s largest companies opt out of the European market, the technological sovereignty of the EU will be threatened. The bloc is certainly not up for that. AI does need a regulatory framework, but this should keep pace with technological advancements. Maybe propose using AI to monitor other AI systems?
In the upcoming series of discussions, we expect to see versions of more acceptable regulations that do not completely throttle innovation. Meanwhile, ChatGPT has written a new law to regulate AI, after being asked by Costa Rica lawyers to draft one according to their constitution!