Google has said it will sign the EU’s code of practice for artificial intelligence models, but criticised the bloc’s AI Act amidst ongoing pressure over the regulations from large tech groups and the US government.
The code, which is voluntary, sets out guidance for how general-purpose AI models should comply with the AI Act, including legal obligations for the safe deployment of the technology, copyright protections and transparency around how the models are trained.
Kent Walker, president of global affairs and chief legal officer at Google parent Alphabet said the company would sign the code in the hope that it will “promote European citizens’ and businesses’ access to secure, first-rate AI tools”.
Corporate opposition
But he said the company would also submit feedback and reiterated previous industry criticism that the AI Act and the code “risk slowing Europe’s development and deployment of AI”.
“Departures from EU copyright law, steps that slow approvals, or requirements that expose trade secrets could chill European model development and deployment, harming Europe’s competitiveness,” Walker stated.
The White House has characterised EU regulations and their accompanying fines as a form of taxation.
In a statement following the signature of a US-EU trade deal on Sunday, the White House said the two economies “intend to address unjustified digital trade barriers”.
The chief executives of large European companies, including Airbus and BNP Paribas, in an open letter asked the bloc for a two-year pause in bringing the regulations into force, saying complexity and a lack of clarity threatened European competitiveness in the industry.
So far the EU has maintained that it plans to continue its schedule for implementing the AI rules in the face of such objections.
‘No pause’
“Let me be as clear as possible, there is no stop the clock. There is no grace period. There is no pause,” said Commission spokesman Thomas Regnier earlier this month.
The implementation schedule includes obligations for general-purpose AI models beginning in August and obligations for high-risk models beginning in August 2026.
The Commission has, however, said it plans to submit proposals for simplifying the rules, such as reducing reporting obligations for small companies, toward the end of this year.
OpenAI and France’s Mistral have agreed to sign the code, while Microsoft told Reuters this month that it was “likely” to do so.
Facebook parent Meta has been the most aggressive in opposing the EU’s attempts to regulate its AI tools, last year delaying the launch of social media AI features in Europe and the release of its multimodal Llama AI model in the bloc over regulatory concerns.
Political clash
It said earlier this month it would not sign the AI code of practice as it “introduces a number of legal uncertainties for model developers, as well as measures which go far beyond the scope of the AI Act”.
Joel Kaplan, Meta’s chief global affairs officer and a political ally of US president Donald Trump, who was appointed to the role just before Trump took office in January, called the EU rules “over-reach” that would “throttle the development and deployment of frontier AI models in Europe, and stunt European companies looking to build businesses on top of them”.
Meta has made aggressive efforts to bring its corporate policies into line with those of the White House since the beginning of Trump’s term, including appointing a personal friend of Trump’s to its board and ending its fact-checking in the US.