Showing how easy it is to make deepfakes of politicians using artificial intelligence, independent senator David Pocock creates AI videos of Prime Minister Anthony Albanese and Opposition Leader Peter Dutton.
Okay, so I have a few points in rebuttal, but I think we’re generally on the same page. I would absolutely support a ban on political parties using deepfakes of opponents in attack ads or otherwise broadcasting them. Infact, I would support more than just fines. If it was discovered after the polls, I would support a full recall of the election. I’d suggest deregistration from that election as well, but I’d hope the voting public can show it’s distaste for that behaviour. I’d also support some level of required due diligence for news media in ensuring what they’re publishing is real. Though there has to be consideration of the suppression of important information. Can you imagine if the Watergate tapes were never released because it couldn’t be proved beyond a shadow of a doubt that they were real?
So I guess that brings me to my problem with this petition. It seems to be asking for a general ban on the entire population, and that’s just not something I can support. There is, and should be, higher standards of ethics expected of both of these groups. However, I don’t think it should be enforced on the average citizen. They just don’t have the ability to get that stuff in front of eyeballs without help.
The other side of this (I know we’d basically be pissing into the wind with how small we are) is that regulations targeting the companies themselves needs to be a part of this. You should never be able to type in the name of a notable figure (or anyone, really) into a generative AI and get it to spit out an image/video of that person. It’s being used to make porn of celebrities which is incredibly damaging, but there’s now been cases of students creating it of other students by feeding it pictures. If AI companies won’t create safeguards, we need to make them. As it stands, it requires an immense amount of power to train picture/video-generating AIs that can fool people. So targeting larger actors makes the most sense.
Your first paragraph about the politics and news media, yeah generally agree with.
The point about ethics in the news media i see as part of the problem surrounding the Australian Press Council and their principle funders being the organisations they supposedly investigate. Its an ethics system set up to generate conflicts of interest. The media have a watch dog that’s more like a chew toy.
We are calling on Minister Farrell to legislate a ban on the use of deepfakes in elections before the next federal election.
From the change.org petition.
Seems that Pocock is referring to AI deepfakes’ use in elections. Maybe i haven’t read the parts your referring to?
While i see no actual wording because i’ve not seen an example of a proposed legislation, I would assume that its wording would be vague enough to potentially catch many people, likely including individual citizens. Aus courts would, as they always try to, sort out the chaff through testing intent of parties accused of using deepfakes. But thats me speculating, and i see nothing other than the quote above to suggest how far they’d go, so “in elections”?.
Your last paragraph is the part i disagree with you on. Its not pissing in the wind to regulate a large company, in fact its a necessity for smaller countries like Australia.
Like you say, targeting these large tech companies makes sense, and like the news media bargaining showed, it can be done. Whatever we think about that particular issue, the tech companies played the governments game. Large countries can find all sorts of excuses to not find consensus, and sometimes need that leading example, and in telecommunications cases it can’t really be a US State that takes the lead, like they can on other issues, but a separate country. I think its due to telecommunications law being federal jurisdiction.
Okay, so I have a few points in rebuttal, but I think we’re generally on the same page. I would absolutely support a ban on political parties using deepfakes of opponents in attack ads or otherwise broadcasting them. Infact, I would support more than just fines. If it was discovered after the polls, I would support a full recall of the election. I’d suggest deregistration from that election as well, but I’d hope the voting public can show it’s distaste for that behaviour. I’d also support some level of required due diligence for news media in ensuring what they’re publishing is real. Though there has to be consideration of the suppression of important information. Can you imagine if the Watergate tapes were never released because it couldn’t be proved beyond a shadow of a doubt that they were real?
So I guess that brings me to my problem with this petition. It seems to be asking for a general ban on the entire population, and that’s just not something I can support. There is, and should be, higher standards of ethics expected of both of these groups. However, I don’t think it should be enforced on the average citizen. They just don’t have the ability to get that stuff in front of eyeballs without help.
The other side of this (I know we’d basically be pissing into the wind with how small we are) is that regulations targeting the companies themselves needs to be a part of this. You should never be able to type in the name of a notable figure (or anyone, really) into a generative AI and get it to spit out an image/video of that person. It’s being used to make porn of celebrities which is incredibly damaging, but there’s now been cases of students creating it of other students by feeding it pictures. If AI companies won’t create safeguards, we need to make them. As it stands, it requires an immense amount of power to train picture/video-generating AIs that can fool people. So targeting larger actors makes the most sense.
Your first paragraph about the politics and news media, yeah generally agree with.
The point about ethics in the news media i see as part of the problem surrounding the Australian Press Council and their principle funders being the organisations they supposedly investigate. Its an ethics system set up to generate conflicts of interest. The media have a watch dog that’s more like a chew toy.
Seems that Pocock is referring to AI deepfakes’ use in elections. Maybe i haven’t read the parts your referring to?
While i see no actual wording because i’ve not seen an example of a proposed legislation, I would assume that its wording would be vague enough to potentially catch many people, likely including individual citizens. Aus courts would, as they always try to, sort out the chaff through testing intent of parties accused of using deepfakes. But thats me speculating, and i see nothing other than the quote above to suggest how far they’d go, so “in elections”?.
Your last paragraph is the part i disagree with you on. Its not pissing in the wind to regulate a large company, in fact its a necessity for smaller countries like Australia.
Like you say, targeting these large tech companies makes sense, and like the news media bargaining showed, it can be done. Whatever we think about that particular issue, the tech companies played the governments game. Large countries can find all sorts of excuses to not find consensus, and sometimes need that leading example, and in telecommunications cases it can’t really be a US State that takes the lead, like they can on other issues, but a separate country. I think its due to telecommunications law being federal jurisdiction.