Google’s President of Global Affairs, Kent Walker, has announced that the company will sign the EU’s General Purpose AI Code of Practice. Google is presenting this as a commitment to promote secure, “first-rate” AI tools that will boost the EU economy by 8% annually by 2034. Google says that the move is a positive step for Europe’s innovation and economic goals.
The Code of Practice is a voluntary agreement that stems from the broader EU AI Act, which is one of the world’s first comprehensive legal frameworks for AI. In its announcement, Google was careful to emphasize the economic benefits of AI for Europe and position itself as a partner to achieve growth, rather than a problem to be dealt with.
Despite Google signing the pact, Google has included a warning that the AI Act and Code “risk slowing Europe’s development and deployment of AI.” Walker also outlined three big concerns that the company had with the AI Act and Code. Google fears departures from EU copyright law, slow approvals, and requirements that “expose trade secrets.”
Exposure of trade secrets is a key concern for all major AI developers, who largely see their model architecture and training data as proprietary. The move by Google to sign the Code is a calculated, qualified agreement. It agrees to play along, but only if the rules don’t harm its business interests.
The search giant has committed to working with the EU"s AI Office to ensure the code is “proportionate and responsive.” Google has also promised to be an active voice for a "pro-innovation" approach, which involves more self-regulation and fewer rules handed down by European politicians.
With this announcement, Google is signaling that it’s not too bothered with AI regulation, but over who defines what “pro-innovation”, “proportionate”, and “responsive” actually mean.
Image via Depositphotos.com