Online cesspits such as X go in only one direction. There is no saving Elon Musk’s platform, and any glimmer of perceived usefulness is far outweighed by the negatives. The scandals will keep rolling. Decency will continue to be shredded, and there are always new depths of vulgarity to be plumbed.
Its latest scandal involves non-consensual abuse material generated at scale by its embedded AI chatbot, Grok. But it was already known that Grok actively encourages this by design. Last year, Grok’s Imagine AI video tool introduced a “Spicy” mode, allowing the creation of sexually suggestive deepfakes, and worse. In one test example published by The Verge last August, when fed a prompt to generate videos of Taylor Swift at Coachella, “several defaulted to ripping off most of her clothing”. This is not an aberration, it is a feature.
And inevitably, such features will be used to torment, harass and demean women and girls for kicks. Grok told X users on Friday that image generation and editing features were now available only to paying subscribers, but that’s hardly going to stem the tide. Inevitably, when you have Musk in charge, guardrails will vanish and chaos will reign.
Politicians need to realise that X is not Twitter. Under Musk, X is a vast disinformation network, a hotbed of racism, hate, extremism and dystopian delusions. It is a radicalisation tool, an arena of harassment, and yes, its chatbot is a creator, publisher and distributor of awful material.
READ MORE
Compared to X, Twitter was a platform more subject to moderation. With X, Musk made a virtue of reducing moderation, while also reinstating banned accounts, especially those previously barred for hateful content. After he acquired what was then Twitter, Musk fired 80 per cent of the engineers working on trust and safety, and let go a third of non-engineering trust and safety employees. The full-time content moderation team was halved, and the number of contracted moderators also fell. Its Trust and Safety Council was disbanded. CNN reported that in the weeks leading up to this scandal, Musk’s xAI safety team “already small compared to its competitors, lost several staffers”.
Using AI to remove real people’s clothes from an existing image, and publishing and sharing that image, is vile. It’s also illegal under the Harassment, Harmful Communications and Related Offences Act 2020. Under Section 5 of the Child Trafficking and Pornography Act 1998, it is illegal to knowingly produce, distribute, print or publish what is referred to in the Act as “child pornography” (child sexual abuse material, CSAM). Also relevant is Section 9 of that Act: where an offence under section 5 is committed by a corporate body, there are also criminal consequences for any officers of that body.
But there is a broader issue here regarding the tone of the political reaction in Ireland last week, which was characterised by obfuscation. There was plenty of shock expressed by politicians, but also a strange reluctance by some to decisively call out the platform.
Among the remarks that felt discordant was Micheál Martin appearing to “both sides” the situation: “Platforms can be misused and abused, or they can be used for positive reasons as well in terms of articulating various positions, Government policies and so on,” he said. “I think one would have to consider those issues carefully before jumping to conclusions in respect of [the] utilisation of various platforms at any given time.” With that, the Taoiseach really was weighing up the importance of promoting the Government’s messaging versus the presence of abusive material.
Peter Burke, the Minister for Enterprise, said Government should make a “collective decision” about whether or not to stay on the platform. Why invent an unnecessary process? X isn’t some kind of international accord or body. There is no obligation or rule that compels any politician, government department or state agency to sign up to it.
[ Are sexualised images being made by Grok AI illegal in Ireland?Opens in new window ]
Perhaps worst of all were Patrick O’Donovan’s remarks, given that he is the Minister with responsibility for media. He said he had deactivated his own X account on Friday morning, but also that “ultimately, at the end of the day, it’s a choice of a person to make these images” ... “It’s not necessarily the app that’s making the images.” O’Donovan is wrong; it is the app making the images.
O’Donovan is not averse to bullish statements when it comes to perceived failings of media companies, as his attitude to RTÉ has shown. Perhaps he should take some of that energy and transfer it to X.
Minister of State for Artificial Intelligence, Niamh Smyth, did at least question the appropriateness of the Government maintaining a presence on the site, and demanding a meeting with X.
There are many reasons why politicians, State agencies, Government departments and legitimate media outlets should break ties with X, something I have previously written about.
There are also arguments for maintaining a presence there, so that a relatively large sphere of social media isn’t completely overrun by toxicity – although as we can tell from the ongoing slide, the toxicity is winning out.
Enforcing the law and regulation at a European level are bigger issues. While indecisiveness and stasis are characteristics of this Government, the reactions of some speak to a curious mode of capitulation to Big Tech, regardless of how rotten a platform and its owner are.
