Дата публикации
Human rights organizations, including EDRi, AccessNow, ARTICLE19, Amnesty International and others, analyzed the AI Act approved by the European Parliament.
The law provides for various measures to protect society: from labeling images created by neural networks without direct human intervention to completely banning some content.
In particular, the use of individual, ethnic or religious characteristics to identify people or recognize their emotions is strictly limited.
While acknowledging progress in accessibility and transparency, NGOs have raised concerns about law enforcement and national security exceptions, as well as partial bans on civil society practices.
In their report, EDRi and colleagues put forward a collective vision of the approach:
“…[an approach] in which ‘human-centred’ is not just a buzzword, but where people are treated with dignity and legislators are bold enough to draw red lines against the inappropriate use of AI systems.”
Human rights activists highlight “many missed opportunities to ensure the protection of rights to privacy, equality, non-discrimination, the presumption of innocence and many other rights and freedoms when it comes to AI.”
According to a number of NGOs, they put a number of demands before legislators, which were either not fulfilled or were satisfied only formally. These are the recommendations:
- Expand the system of accountability, transparency, accessibility and redress for people harmed by AI systems;
- Limit discriminatory surveillance by law enforcement and migration authorities;
- Fight back against the lobbying of big technology companies.
The AI Act does not meet a number of civil society demands, human rights activists believe.
The law may be finalized in its final political and technical form as early as April or May of this year, and some of its provisions will come into effect as early as November.
“It is therefore vital that civil society groups are given a seat at the table and that this work is not carried out in opaque conditions. We urge legislators around the world who are also considering introducing horizontal rules on AI to learn from the many mistakes of the EU outlined above,” the NGO press release concludes.
(translation was carried out automatically)
The law provides for various measures to protect society: from labeling images created by neural networks without direct human intervention to completely banning some content.
In particular, the use of individual, ethnic or religious characteristics to identify people or recognize their emotions is strictly limited.
While acknowledging progress in accessibility and transparency, NGOs have raised concerns about law enforcement and national security exceptions, as well as partial bans on civil society practices.
In their report, EDRi and colleagues put forward a collective vision of the approach:
“…[an approach] in which ‘human-centred’ is not just a buzzword, but where people are treated with dignity and legislators are bold enough to draw red lines against the inappropriate use of AI systems.”
Human rights activists highlight “many missed opportunities to ensure the protection of rights to privacy, equality, non-discrimination, the presumption of innocence and many other rights and freedoms when it comes to AI.”
According to a number of NGOs, they put a number of demands before legislators, which were either not fulfilled or were satisfied only formally. These are the recommendations:
- Expand the system of accountability, transparency, accessibility and redress for people harmed by AI systems;
- Limit discriminatory surveillance by law enforcement and migration authorities;
- Fight back against the lobbying of big technology companies.
The AI Act does not meet a number of civil society demands, human rights activists believe.
The law may be finalized in its final political and technical form as early as April or May of this year, and some of its provisions will come into effect as early as November.
“It is therefore vital that civil society groups are given a seat at the table and that this work is not carried out in opaque conditions. We urge legislators around the world who are also considering introducing horizontal rules on AI to learn from the many mistakes of the EU outlined above,” the NGO press release concludes.
(translation was carried out automatically)
