That’s the argument in an open letter and newspaper ad signed by 49 corporate executives, researchers, and industry groups, including the CEOs of SAP, Spotify, and Ericsson. But the people who framed the entire communication, which appears to be Meta’s idea, are Meta CEO Mark Zuckerberg and chief AI scientist Yann LeCun.
“We are essential businesses, researchers and institutions for Europe, working to serve hundreds of millions of Europeans,” the letter begins, arguing that “if businesses and institutions are to invest hundreds of billions of euros to build generative AI for European citizens, they need clear, consistently applied rules to enable them to use European data.”
But unfortunately, he points out, “regulatory decision-making has become fragmented and unpredictable in recent years, and the intervention of European data protection authorities has increased uncertainty about what kind of data can be used to train AI models.” The EU thinks it is protecting citizens from the risks of AI development, but in reality it is creating confusion by layering different regulations, he explains.
These regulations are centered around two pieces of legislation: the recently enacted AI Act and the GDPR, which enforces data privacy regulations. Of course, the former is not mentioned in the letter, and the latter, GDPR, is mentioned only in passing. So the core of the complaint is that, no matter how well-intentioned and carefully written the regulations are, the application of these regulations is harming technology companies that are trying to do great things with AI.
Many people worry that there is too little regulation of AI, and that by the time the world wakes up, it will be years too late to prevent the worst excesses.
This open letter takes the opposite view: too many of the wrong kinds of rules that don’t understand how AI works are holding back progress in the field.
Meta’s fingerprints on the open letter
In June, Meta told users it planned to train AI models using Facebook and Instagram data, sparking controversy with pushback from regulators in the EU and the UK, and privacy groups accusing it of violating GDPR. Facebook has since relaunched its AI data program in the UK, but has been prevented from doing so in the EU.
This EU regulation appears to be the backdrop to this open letter. Several companies signed it, but Meta’s fingerprints are on it, having hosted the letter on its servers and paid for advertising in the Financial Times.
The body of the letter also mentions Facebook’s Llama LLM. “Frontier, open models like Llama, based on text or multi-modal, could boost productivity, accelerate scientific research and add hundreds of billions of euros to the European economy,” it said. But it warned that “without them, AI development will take place elsewhere, depriving Europe of the technological advances enjoyed by the US, China and India.”
This sentence, along with the hosting server, suggests that this open letter was drafted by Meta with the approval of the other signatories. This is bound to lead to a cynical reaction that what appears to be an open letter is actually a thinly disguised PR campaign by Meta to assert its position in the fight against the EU.
Nick Clegg, Meta’s President of Global Affairs, tweeted his support for the letter: “Today, a number of leading European businesses, researchers and developers are urging the EU to adopt a streamlined approach to data regulation or risk falling behind in AI innovation: http://EUneedsAI.com.”
Inseed AI Community co-founder Robert Makieko, a fierce critic of Meta’s approach, responded to Clegg’s tweet with a curt response:
“It’s really sad that you’re participating in an Orwellian double standard. Mark Zuckerberg wants the right to use your data and property for his own benefit forever, without asking or compensating you. By ‘user’ I mean anyone in the world, whether an individual, a company, a brand, or a country. If his clones replace your work, that’s your problem, not his.”
The real depth of the objections and concerns
The counterargument to this open letter is that regulations like the EU AI law do not make life harder for AI developers, but rather provide a firm foothold for companies to make future decisions. In other regions, much remains unknown.
Of course, we can’t ignore the concerns of the other signatories to the open letter, but it’s hard to know exactly how deep they are. Computerworld contacted several of the signers, and received the following response from SAP:
“SAP encourages policymakers to adopt a risk-based, outcome-oriented approach to AI policy and develop a legal framework that builds on existing law and does not duplicate or create conflicting requirements. SAP’s ethos is deeply rooted in European values and is focused on bridging Europe’s innovation gap in a responsible way that protects the well-being of its citizens.”
This avoids directly criticizing the EU’s current AI regulations or piggybacking on calls for freedom for private companies.
Any group or individual who agrees with the contents of the open letter can add their signature by filling out a form that sends details to Meta to “join the call for certainty in AI regulation in the European Union.” It is unclear how these “signatures” will be verified.
editor@itworld.co.kr
Source: www.itworld.co.kr
