This is a good way to regulate this. Criminalize people who abuse AI tools to cause harm. Don't try to impose censorship or mass-surveillance on AI tools. I oppose all pornography, but censoring nudity from a model both compromises the model's quality (example: SD3) and stops legitimate artistic value.
Though the framing on Grok is highly duplicitous. It is against the ToS, and it's about a few abusers among millions of legitimate users. Meanwhile there are actual "nudification" services which advertise themselves entirely to enable this kind of abuse.
How does the law distinguish itself from anti-harassment laws?
I recall a old lawsuit (not sure which country) where a student had photoshopped a nude body onto a picture of their teacher and given it to other students. The main question in the case was if this qualified as sexual harassment, even if the teacher in question never received the image. I don't remember the outcome, through I think they were found guilty.
"Ireland fast tracks Bill to criminalise harmful voice or image misuse" - now that's some strange capitalis/zation in that headline. I'm Ok with title case, but randomly capitalising just one word in a title looks strange. Or is this something specifically Irish?
> knowingly uses or infringes upon the use of and publishes ... an individual’s name, photograph ... without the individual’s prior consent ... and being reckless as to whether or not harm is caused to, the other person.
> [harm occurs when someone] seriously interferes with the other person’s peace and privacy or causes alarm or distress to the other person
This seems very widely worded. A newspaper publishing the name/image of a suspected criminal is definitely "publishing an individuals name, photograph", without their consent, and can quite clearly cause alarm or distress.
Without some exemption clauses added, this bill seems to basically ban using anyone's name/photograph/likeness in ANY context that criticises them; it will almost certainly conflict with ECHR's Article 10 on freedom of expression. However(!!) with a few exemptions it can be made much better. Even tying it to AI generated photos/voice/etc would help - most _genuine_ criticism and reporting can go without the use of AI, but a lot of the intentional harm and sexual harassment did not occur before AI. If they don't want to do that, adding some form of "exemption if the information was used in a non-libellous context" could also work.
> The deliberate misuse of someone’s image or voice without their consent for malign purposes should be a criminal offence.
So much vagueness in misuse yet so much impact with consequences. This bill is being speedlined. And it seems to hurt everyone at the same time.
Better to bring a bill with specific limitations and prohibitions that is complete with treatment, consequences, prosecutions, and appeals.
Then that bill can be extended to other prohibited actions as they become apparent
This is the entire proposed bill:
https://www.oireachtas.ie/en/bills/bill/2025/11/
https://data.oireachtas.ie/ie/oireachtas/bill/2025/11/eng/in...
Interpretation
1. In this Act—
“broadcast” has the meaning assigned to it by the Broadcasting Act 2009;
“distribute” means distribute to the public or a section of the public;
“harm” includes psychological harm;
“Minister” means the Minister for Justice;
“publish” means publish, other than by way of broadcast, to the public or to a portion of the public.
Offences
2. (1) A person who—
(a) knowingly uses or infringes upon the use of and publishes, performs, distributes, transmits, or otherwise makes available to the public an individual’s name, photograph, voice, or likeness in any medium for purposes of advertising products, events, political activities, merchandise, goods, or services or for purposes of fundraising, solicitation of donations, purchases of products, merchandise, goods, or services or to influence elections or referenda, or
(b) distributes, transmits, or otherwise makes available an algorithm, software, tool, or other technology, service, or device, the primary purpose or function of which is the production of an individual’s photograph, voice, or likeness, is guilty of an offence where these actions are carried out—
(i) without the individual’s prior consent, or, in the case of a minor, the prior consent of such minor's parent or legal guardian, or in the case of a deceased individual, the consent of the executor or administrator, heirs, or devisees of such deceased individual,
(ii) with intent to cause harm to, or being reckless as to whether or not harm is caused to, the other person.
(2) For the purposes of subsection (1), a person causes harm to another person where—
(a) he or she, by his or her acts, intentionally or recklessly seriously interferes with the other person’s peace and privacy or causes alarm or distress to the other person, and
(b) his or her acts are such that a reasonable person would realise that the acts would seriously interfere with the other person’s peace and privacy or cause alarm or distress to the other person.
(3) A person who is guilty of an offence under this section is liable—
(a) on summary conviction to a class A fine or imprisonment for a term not exceeding 12 months, or both, or
(b) on conviction on indictment to a fine or imprisonment for a term not exceeding seven years, or both.
Review and operation of Act
3. The Minister shall, not later than one year after the commencement of this Act, carry out a review of the operation of this Act.
Short title and commencement
4. (1) This Act may be cited as the Protection of Voice and Image Act 2025.
(2) This Act shall come into operation on such day or days as the Minister for Justice may by order or orders appoint either generally or with reference to any particular purpose or provision and different days may be so appointed for different purposes or different provisions.AI is a god send to governments who have been trying to push through censorship laws under the guise of "protecting [people|children|women|etc]"
The UK is considering an injunction to have Grok and potentially X banned until this issue is resolved.
I'm no huge fan of state intervention, particularly my state which is notoriously over zealous, prudish, and subtly authoritarian. The sweeping changes were seeing in the UK are somewhat reminiscent of the old "First they came for the trade unionists" poem, and there certainly has been many of us speaking out against the attacks on freedom to protest and expression here.
However, when you swap out trade unionists for "People comitting sexual harassment", "Paedophiles", and "Billionaires", Im somewhat more inclined to side with the government on this. Were already way down the slippery slope, at least some genuinely bad actors are catching some flak this time I guess.
Sure, its a proposed law for citizens of this country.
So, how's that going to work out with Indian, Pakistani, North Korean, or similar locations using voice spoofing?
Thats right. None.
> Ireland fast tracks Bill
Poor Bill, what did he do to deserve this?
Ireland is just a copy of the UK, except it's a tax haven and a trojan horse of the EU. They have the same age requirement regulation online as the UK.
So the actual context of this is a long-standing feud between Meta and our elected representatives (and some prominent media figures) regarding social engineering scam ads.
You've probably seen some variant of it - a 1:1 copy of a reputable newsoutlet - in this case The Irish Independent, The Irish Times, RTÉ or Newstalk - with a purported video from our Prime Minister or similar which leads to a crypto-trading or forex-trading sales funnel. Simon Harris, our previous Head of State, was inundated with variants of the same script:
“In the nation’s best interests, we’ve carried out a full investigation to make sure it’s not a scam,” the AI-generated Harris appears to say in the scam ad shown on YouTube.
“This is your chance to change your life. All it takes is one small step: invest €250 and start earning today,” the deepfaked figure tells a press conference.
https://www.thejournal.ie/facetcheck-debunk-ai-scam-ad-deepf...
https://www.newstalk.com/news/social-media-platforms-see-sur...