Reporters Without Borders (RSF) has
abandoned work on drafting the code of practice for general
purpose artificial intelligence models, such as OpenA's Gpt-4,
arguing that the latest version is "largely insufficient" and
denouncing "the lack of guarantees on the right to information
and the exorbitant weight of industry in this process".
In recent days, the third draft of the code, required by
European law on AI, has sparked protests from industry
professionals, including the architects who negotiated the text
in the European Parliament and the EU Council.
On Thursday the organisation protecting journalists and freedom
of the press decided to abandon the work.
According to RSF, "there is a lack of attention to the
protection of reliable information: the right to information is
not mentioned in the text".
Also absent, the NGO complains, are "the risks associated with
the unregulated development of artificial intelligence for
reliable information, such as deepfakes, the proliferation of
automated fake news sites or even disinformation infiltrated by
chatbots".
RSF finally denounces the fact that "fundamental rights, such as
systemic risks to the integrity of democratic elections, are
relegated to an appendix and their consideration is optional".
The code of good practices, whose final draft is expected in
May, will have to detail the rules of the AI ;;Act that apply to
providers of AI models for general purposes, in particular those
with systemic risks.
ALL RIGHTS RESERVED © Copyright ANSA
