Mike Schentrup’s Post

The continued development of A.I. models have led to new ethical issues that were barely imaginable just a handful of years ago, like the ease and speed at which deepfake porn videos and images can now be created. Existing laws and jurisprudence has not been able to fill the gaps, so the government is stepping in to draft new laws to combat the misuse of this technology: "The Senate unanimously passed a bipartisan bill to provide recourse to victims of porn deepfakes — or sexually-explicit, non-consensual images created with artificial intelligence... 'Current laws don’t apply to deepfakes, leaving women and girls who suffer from this image-based sexual abuse without a legal remedy,' Durbin posted on X after the bill’s passage. 'It’s time to give victims their day in court and the tools they need to fight back.'" Read more: https://lnkd.in/eZH95Mgu . . . . #attorney #lawyer #law #lawfirm #nyclaw #nyclawyer #legal #attorney #litigation #attorneylife #advocate #socialmedia #mediaproduction #marketing #socialmediamarketing #contentcreator #AI #openAI #articifialintelligence #AIgenerated #AIgeneratedart #AIart #deepfake #government #AOC

So, WOW. Since WHEN is it OK for the FBI to outright lie about the law to the American public? I am no supporter of CSAM. I AM a supporter of Supreme Court precedent and the notion that just because the FBI may want something to be illegal does not make that so. The FBI issued a "warning" to the public that AI generated CSAM is child pornography (subject to mandatory minimums at the federal level and in many states). HOW ODD, given that the SCOTUS issued a decision in Ashcroft v Free Speech Coalition, 535 US 234 (2002) (link to case in comments)-- that's 22 years that this has been the LAW OF THE LAND, folks, explicitly holding that computer generated CSAM is protected by the First Amendment. "The freedom of speech has its limits; it does not embrace certain categories of speech, including defamation, incitement, obscenity, and pornography produced with real children." pp. 235-246. "Ferber upheld a prohibition on the distribution and sale of child pornography, as well as its production, because these acts were "intrinsically related" to the sexual abuse of children in two ways [Ferber link also in comments]. First, as a permanent record of a child's abuse, the continued circulation itself would harm the child who had participated. Like a defamatory statement, each new publication of the speech would cause new injury to the child's reputation and emotional well-being. . . Second, because the traffic in child pornography was an economic motive for its production, the State had an interest in closing the distribution network." "In sum, § 2256(8)(B) covers materials beyond the categories recognized in Ferber and Miller, and the reasons the Government offers in support of limiting the freedom of speech have no justification in our precedents or in the law of the First Amendment. The provision abridges the freedom to engage in a substantial amount of lawful speech. For this reason, it is overbroad and unconstitutional." Sec 2256(8)(B) was the section of the CPPA that prohibited computer generated images (AI) depicting sexual activity of minors. Furthermore, the cases the FBI did cite should be overturned on appeal under the Ashcroft v Free Speech holding. It is unconscionable that the FBI and the USDOJ would bring forward cases alleging and ultimately convicting accuseds of acts that are clearly constitutionally protected conduct under established Supreme Court precendent that is still good law. If they don't like the constitution, change it, or as some people say, move to another country. https://lnkd.in/eCxRwVxP

1/ The UK government plans to introduce new legislation that will criminalize the creation and distribution of sexually explicit deepfakes, closing a legal loophole that has left victims without recourse. 2/ While publishing intimate photos or videos without consent has been illegal in the UK since 2015, there are currently no laws specifically targeting AI-generated fakes, prompting the government to take action. 3/ The problem extends to schools, with 60% of teachers concerned that their students could be involved in deepfake scandals, despite 73% of parents believing their children are not involved. In Florida, two teenagers were recently arrested for creating and sharing AI-generated nude images of their classmates.

AI is confusing the law on CSAM and pornography, challenging its boundaries. A few genuine questions about AI CSAM involving known celebrities and AI teen/barely legal material. How do we categorise the following scenarios and what does this mean for regulation? ➡ AI is used to generate what looks like a child engaged in sexual activity but the face/likeness is of a known person/celebrity over 18. - Is this CSAM as the body/sexual activity is of a child, even tho we know its a person over 18? Possibly not, as we know they're over 18. - But what if AI also used to make the celebrity look like a child? We still know it's X celebrity, but now they look like a child. Possibly yes? It's imagery of a child, and if you didn't know X celebrity then you would think it's CSAM? - English law says its a prohibited image of a child if 'the impression conveyed by the image is that the person shown is a child'. - Not sure this necessarily clarifies things: it looks like a child, but we know it's not as it's X celebrity. Is it reliant on how closely it resembles X celebrity or other known individual? ➡ Then, what of 'teen' and 'barely legal' imagery generated by AI but which is very similar to that involving actors known to be over 18? - There is a lot of material on mainstream porn websites classified as teen, or labelled 'barely legal' where the actor is over 18 but the image conveyed is of a young child, sometimes very young looking. This is lawful pornography (though I think this is problematic.) - But we could use AI to generate an almost identical image/video and it would be CSAM as it would be classed as sexual activity involving a child. - So, now we have the strange situation where the AI generated image is unlawful, but the one involving a real actor is legitimate. Yet, they both convey children involved in sexual activity; both encourage and legitimise sexual arousal from images of child sexual abuse. - If the AI generated image is unlawful, why not the so-called lawful teen/barely legal images? - And, surely soon enough, the material on porn sites classed as teen could some/mostly/all be AI generated? Implications for law and regulation? - The criminal laws on CSAM and porn could be clarified, and for the millions viewing and using teen/barely legal porn this would be a wake-up call. But realistically, that would not make much of a difference in practice, as law enforcement is already struggling to deal with other forms of CSAM. - But these challenges raise significant issues for regulation right now in terms of platforms removing illegal material under legislation such as the UK's Online Safety Act and the EU's Digital Services Act. Is AI teen/barely legal material illegal? Any thoughts very welcome. For more on AI and CSAM, see Internet Watch Foundation (IWF) Michael Tunks https://lnkd.in/etAEh-YM

There's another solution to the horrific issue Robin Tombs highlights - FUND FEMALE AI FOUNDERS. Read my Fast Company op-ed '3 Reasons The Future Of AI Relies On Women': https://lnkd.in/gF9JGCUX Last year, only 1.7% of all venture capital went to female founders - even less goes to female AI founders. Fund founders like Melissa Hutchins, building the solution to this at Certifi AI #FundFemaleFounders

NH bill targets deepfake sexual images, adding to 'revenge porn' law. Here's why. LINK: https://lnkd.in/ekVcne8B Hashtags: #TheLegalLowdown #TheLawandTechnology #Law #Technology Please LIKE & REPOST

🤔 An *interesting* read 👇 on the #criminalisation of so called (in the article) #deepfake pornography. While I disagree with most of it (such as - there is a BIG difference between having sexual fantasies in your mind and creating and distributing sexual images/videos of women they didn’t consent to being in), it highlights the importance of #nuance in understanding and redressing harms facilitated by #technology. I think a binary perspectives when seeking to understand technology-facilitated abuse and violence is problematic, not least the implication of so called deepfake videos causing less harm to the women featured than other forms of Image-Based Sexual Abuse #IBSA as they are “fake”. We don’t know this to be true and the limited existing literature suggests otherwise (#MyBodyMyChoice). Placing abuse in a hierarchy of harm continues the harm caused as victim-survivors are confronted with victim-blaming and minimising attitudes in society. As #AI tools continue to advance at such a rapid rate without widespread regulation, the distinctions between offline/online, real/fake become increasingly blurred boundaries it is no longer helpful to maintain! What do you #think? 🤔

Another news article today covering non consented publication of #deepfake #porn. Jonathan Bates superimposed the heads of real women on to naked images carrying out sex acts. Sunday Times journalist Louise Eccles explains that publishing pornographic images online without the subject’s permission was outlawed by the Online Safety Act last year but it is still legal to create deepfakes. Eccles notes that the safeguarding minister, Jess Phillips has said the govt will introduce legislation to crack down on false pornographic images. The victim in this case explains that when she she was trying to get some of the websites to listen and help take down the images, she couldn’t get anywhere. It is now illegal for online platforms to display pornographic images without the subject’s permission, accessible to UK located viewers. If non consented image abuse continues on many platforms (which it most likely will) I think we will begin to see in the UK: 1) a growing number of civil & criminal cases against platforms that choose to continue to display intimate images where faces are clearly visible but no eSign consent has been captured from the person matching that face. 2) given the seriousness of this online harm, increasing pressure on Ofcom to require bigger (& not unreasonably smaller platforms) to use preventative #safetytech to avoid the risk of displaying unconsented naked content. It costs less that £0.50 to use Yoti tech (eSigning + facial auth + face matching) to gain clear fce matched consent from the one or two subjects in an image or video. It’s hard to see why a UK Court would let off the owners / directors of platforms who say they didn’t want to pay that level of fee to use safetytech to ensure they can monetise performer consented porn. No consent, no publication.

Progress! IBM is proud to support the TAKE IT DOWN Act, introduced today. Our statement of support and a link to learn more about this landmark legislation is below. “It is profoundly concerning when bad actors misuse AI to create deepfakes of other individuals without their consent, and it is particularly abhorrent when it’s done to create nonconsensual intimate images and pornography. IBM applauds Senators Ted Cruz and Amy Klobuchar for introducing legislation that would create strong criminal and civil liability for people who distribute nonconsensual intimate audiovisual content, including AI-generated content, as well as for people who threaten to do so. Solving the problems posed by deepfakes will require thoughtful, whole-of-society approaches leveraging both changes in law and technology.” – Christina Montgomery, Chief Privacy & Trust Officer, IBM https://lnkd.in/egpTvNVw

About 🕰️! 🇮🇳 let’s learn! We have an opportunity with the #DigitalIndiaAct to curb the menace of sexually explicit #deepfakes! Creation, distribution, storage! We need to have a nuanced conversation, do research and come up with a strong law that can be implemented.