On Friday 18 July, a week after the publication of the long-awaited ‘General-Purpose AI Code of Practice’ (see EUROPE 13678/17), the European Commission published ‘guidelines’ to help suppliers of AI models comply with the obligations of the AI Act, some of which will come into force on 2 August.
These guidelines should clarify these obligations, in particular by defining the types of models concerned, what is meant by “supplier” and “placing on the market”, and the criteria for certain exemptions.
They also explain the implications of adhering to the ‘Code of Practice’ and describe the Commission’s expectations in terms of compliance, as well as the interpretation and application of the AI Act as understood by the institution.
These guidelines were requested by many players in the sector, who were quick to criticise the short time remaining between the publication of the ‘Code of Practice’ and the entry into force of the obligations of the AI Act in August (see EUROPE 13673/15).
The Commission hopes that these guidelines can provide clarification and legal certainty, which have been lacking until now, for the AI providers concerned.
In the afternoon, however, Meta’s head of global affairs, Joel Kaplan, said in a post on LinkedIn that Meta would not be signing the ‘Code of Practice’ (see EUROPE 13679/7).
“Europe is on the wrong track when it comes to AI. We have carefully examined the European Commission’s ‘Code of Practice’ [...] and Meta will not be signing it. This code introduces a number of legal uncertainties for model developers as well as measures that go well beyond the scope of AI legislation”, he said.
To see the guidelines, go to https://aeur.eu/f/hxv (Original version in French by Isalia Stieffatre)