Feedback on draft possible up to 30 March, before the final Code of Practice on General-Purpose AI is set to come out in May.
ADVERTISEMENTThe tech sector remains concerned about a proposed set of rules for providers of General-Purpose Artificial Intelligence (GPAI) after the latest draft was published by a European Commission- appointed expert group on Tuesday, several lobby groups told Euronews.
The Code of Practice on GPAI should help providers of AI models – tools that can perform many tasks such as ChatGPT, Google Gemini and picture application Midjourney – comply with the EU’s AI Act and includes transparency and copyright-related rules, risk assessment, and mitigation measures.
Previous versions of the text have raised copyright issues, among others, and industry representatives such as publishers and rights-holders are dissatisfied with the latest updates, sector operators told Euronews.
Boniface de Champris, senior policy manager at tech lobby group CCIA said that “serious issues remain”, including “far-ranging obligations regarding copyright and transparency, which would threaten trade secrets, as well as burdensome external risk assessments.”
“The new draft makes limited progress from its highly problematic predecessor, yet the GPAI Code continues to fall short of providing companies with the legal certainty that’s needed to drive AI innovation in Europe,” he added.
Elias Papadopoulos, director of policy at internet lobby group DOT Europe, said that at first glance the draft “has been somewhat improved”, but that some provisions still go beyond the requirements of the AI Act.
“For example, mandatory third-party risk assessment pre- and post-deployment, although not an obligation in the AI Act itself, unfortunately remains in the new draft,” he said.
The expert group, which includes actors from the EU, US and Canada, last month pushed back a previous deadline to reflect the feedback of stakeholders it received. Some 1000 participants have attended plenary sessions and workshops designed to help develop the Code since the kick-off in September.
Feedback on this third version can be submitted until 30 March.
The final version is now set to come out in May at the earliest, Euronews reported last month.
Copyright issues
The concerns were echoed by Iacob Gammeltoft, senior policy manager at News Media Europe – which represents 2700 news brands, online and in print, on radio and tv.
“Unfortunately, the latest draft raises serious questions about whether no code is better than this code. It reads like the chairmen got lost in the exceptions of copyright law, and forgot to assess how the general principles of copyright should apply,” said Gammeltoft.
In January, a group of 15 different European rightsholder organisations warned the Commission that the current draft of the CoP contradicts copyright law.
News Media Europe said the “same problems” exist today as in the first draft. “Copyright law creates an obligation of results which requires that lawful access is achieved, and it’s just not good enough to ask AI companies to make ‘best efforts’ to not use our content without authorisation,” he said.
From a fundamental rights’ perspective, the document has also been weakened, said Laura Lazaro Cabrera, programme director for equity and data at the Centre for Democracy & Technology Europe.
ADVERTISEMENT“The third draft confirms what many of us had feared – that consideration and mitigation of the most serious fundamental rights risks would remain optional for general-purpose AI model providers,” she said.
The EU executive can decide to formalisethe CoP under the AI Act – which will be fully applicable in 2027 – through an implementing act.
The issue of Copyright and AI is also subject of European Parliament scrutiny: German lawmaker Axel Voss (EPP) will work on an own-initiative report on the issue.
