Social network due to come before Dail committee on 4 February
Life
Grok, the AI chatbot used in the Elon Musk-owned social network X, is having its photo editing capabilities limited again following uproar over the creation of sexualised images of women and children.
Controversy erupted after pictures of women were posted to X which were then altered by Grok to show them in revealing clothing or swimwear.
X responded by limiting image editing capabilities to paying subscribers. This has been followed up by geoblocking measures preventing its use in certain countries.
Musk has defended Grok’s NSFW (not safe for work) setting, saying it allowed for upper body nudity only, in line with film classification norms in the US, specifically the R-rating afforded to many big budget films.
A statement released on X read: “We remain committed to making X a safe platform for everyone and continue to have zero tolerance for any forms of child sexual exploitation, non-consensual nudity, and unwanted sexual content… We have implemented technological measures to prevent the Grok account from allowing the editing of images of real people in revealing clothing.”
The response by public representatives has been scathing.
A statement from Labour Party TD and Chairperson of the Oireachtas Arts, Media, Communications, Culture and Sport Committee Alan Kelly read: “It is unbelievable that it has taken X as long as it has to make a statement regarding the use of its AI. To reproduce images in a sexual or nudified manner is an utter violation of a person’s right, it is appalling to hear the stories of those who have come forward to share their experience. It should not have happened, and it can never be allowed to happen again.
“There has been enormous backlash against it, an exodus of people from the platform and countries like Indonesia and Malaysia have moved to ban the platform… When it comes to sexualised or nudified images it’s clear that moderation alone simply will not work. AI tools that edit and produce such images should simply not be allowed.”
A representative from X is expected to go before the Committee on 4 February.
At time of writing some 200 investigations into the creation and sharing of explicit images through Grok were being investigated by An Garda Siochana.
What can be done now?
As government scrambles to find appropriate measures to deal with the Grok scandal and avoid a repeat across other very large online platforms, there exists a series of measures open to law enforcement agencies and the State.
According to a joint statement issued by the Irish Council for Civil Liberties and Digital Rights Ireland An Garda Síochána, Coimisiún na Meán, the EU Commission and the Data Protection Commission already have investigatory powers.
The creation of nudified images of children and distribution of intimate images without consent fall under the Child Trafficking and Pornography Act, 1998, and Harassment, Harmful Communications and Related Offences Act 2020 (also known as Coco’s Law) respectively.
From the perspective of Coimisiún na Meán, creating and sharing deepfakes contravenes the Online Safety and Media Regulation Act 2022 and the European Commission also has standing under the Digital Services Act (DSA). X was fined €120 million last month under the DSA over the blue check mark designating a verified identity, as oposed to a paid subscriber; a failure to provide researchers access to data; and a lack of transparency about its advertising practices.
The use of Grok also forms a test of the EU AI Act, which prohibits threats to safety and livelihood by “manipulative AI”. If found to be in breach X could be fined €35 million or 7% of global turnover, whichever is greater.
TechCentral Reporters