The AI era has dawned, and the impact of this technology is beginning to be felt across the adult industry. We are already seeing a plethora of content, synthetic interactions and customizable avatars enabled by artificial intelligence. However, it is clear that the ultimate potential of artificial intelligence in adult remains to be tapped.
The focus of this report is therefore squarely on the future. Will it resemble the Wild West days of the early internet, with disruptive tech shooting up the joint and the market bucking like a wild bronco? Will generative AI, which can produce text, audio, images and more, fulfill users’ ultimate desires with the most customizable, responsive and hyperrealistic experience possible? Or will it devolve into a quagmire of intellectual property theft, consent-based issues over image and likeness, run-ins with the FTC and major card brands, and an arms race of technological sophistication?
Mitigating risk and reassuring banks will require proactive steps like showcasing content moderation policies.
To understand the payment, legal and content ramifications of the coming AI tsunami, XBIZ reached out to an array of industry thought leaders for their perspectives on everything from merchant compliance to new business opportunities sparked by AI.
Payment Processing in the Age of AI
Paysite and platform operators interested in tapping into AI will need to stay abreast of pressing payment processing considerations while developing their AI initiatives.
Generative AI models, while currently costly, are poised to become more accessible as the technology matures, notes Mitch Platt of Vendo.
“When you hear about AI in adult, one of the first things that comes to mind is generative AI,” he says. “Particularly its ability to create video content without the usual requirements such as creators, producers or studio sets. This is indeed a powerful use of this technology, although it’s not entirely new. However, that is just one application among many.”
In addition to being widely used for image generation and modification, Platt points out, generative AI can be employed in developing customer support chatbots, creating marketing campaigns, automating documentation processes, generating reports and aiding in code development. Additionally, it can facilitate employees’ access to information by answering simple questions, thereby eliminating lengthy wait times for interdepartmental information requests at larger companies. The first step for merchants is therefore determining where AI fits within the roadmap of their business.
“In the fintech industry, AI serves multiple purposes, including optimizing the customer journey through the site, refining pricing models, conducting tests on checkout pages and implementing algorithms for fraud prevention,” Platt continues. “The latter is particularly crucial in high-risk industries. Fraudsters are leveraging AI just as much as legitimate businesses, making it imperative to continually refine strategies to stay ahead.”
As for content generation, Platt strongly emphasizes compliance with regulatory standards, particularly regarding consent and harm prevention.
“Mastercard’s standards state that consent must be established if AI images are based on a specific person,” he explains. “If there’s a likeness to a person, standard practices apply — 2257, model agreement, proof of age and so on. If there’s doubt about consent, the content is considered to be violating Mastercard standards.
“The EU AI Act is the first comprehensive regulation of AI,” Platt notes. “Among its core principles are: Document how your AI systems work. Keep your terms and conditions updated with the capabilities of your AI tools. Don’t manipulate, meaning don’t make people do something they wouldn’t do. If you are using a performer as a ‘base model,’ make sure you have a contract specifying what content can be autogenerated based on that performer, and for how long.”
Ensuring adherence to those principles, he notes, requires having strong content removal protocols, such as the ability to monitor content and comply with takedown requests within 24 hours.
“Keep up with regulation,” Platt counsels. “Get legal advice if needed. While generative AI holds immense potential, it remains a dynamic field with constant ongoing developments.”
Cathy Beardsley of Segpay sees that dynamism as a source of uncertainty for some players.
“Just the sheer unknown nature of AI makes people uneasy,” she says. “Several banks we work with have been hesitant to accept or work with AI merchants and AI websites. They say they’re just not comfortable yet with the technology and all that surrounds it. Banks will also conduct a longer review process on an AI website, so that they fully understand the business model and any risk associated with AI.”
Mitigating risk and reassuring banks, Beardsley adds, will require proactive steps like showcasing content moderation policies.
“This will help banks better understand your business model,” she says. “AI-generated content needs to be treated the same way as user-generated content: It must adhere to all the same card brand content polices, like no underage models, bestiality, forced content, etc.”
AI websites, Beardsley says, need to have takedown request links in case users feel that certain AI content is problematic or resembles their likeness, and they need to establish that deepfakes — such as AI images depicting celebrities naked — are not allowed.
“This is a fineable offense which can come with even more serious consequences,” she cautions.
“If the website has content the consumer can modify, you need to ask: Does the website owner have the license for that content and the authorization to modify it?” Beardsley continues. “Lastly, if you’re using AI to moderate your content, you need to know how accurate it is. While it might be easy for AI to determine model age, how about recognizing some prohibited content categories such as bestiality or force/rape? These are really important things to be aware of, not only for the sake of compliance but also simply to help prevent them from being online.”
MobiusPay’s Jonathan Corona agrees with Beardsley’s assessment of banks’ concerns and the key steps for addressing them.
“The card schemes are not ardently opposed to AI-generated content, but they approach the topic with apprehension because proper protocols may not be in place,” he says. “Anything that could be construed as brand-damaging for them is prohibited.”
Merchants who use AI to scan and monitor content, Corona adds, should take extra care and also manually review content to ensure that the AI scanner is not inadvertently approving prohibited content or removing permissible content.
As for AI-generated text, such as sexting, Corona underscores that this also falls under best practices guidelines, and that chatbots should not be able to be prompted to describe anything illegal or brand-violating.
“Many merchants who are engaged in AI chat and image-generation business models have implemented the same filters and safety checks that are used for traditional content,” he says. “Consent and proof of age must be documented, just like any other production.”
Even if a person is generating a likeness of themself, proper documentation is still required in order to maintain compliance with 2257 and card scheme regulations, he emphasizes. Ultimately, it is the merchant’s responsibility to ensure that content is legal in all jurisdictions and compliant with card scheme best practices.
“Just like with non-AI content, complaints and takedown requests should be addressed and resolved within seven days, and records of them should be maintained by the merchant and available for inspection by the processor, sponsor bank or card brands at any time,” Corona adds.
For some, uncertainty and caution become prohibitive.
“Some platforms don’t allow the uploading of any AI images, to ensure compliance with the consent requirement,” Corona points out.
Given the quicksilver pace of advancements in AI tech, payment processors are closely watching regulatory developments.
“General Data Protection Regulation (GDPR) compliance in the EU already poses challenges for AI development, particularly the requirement to explain algorithmic decisions,” Platt says. “For instance, if you are developing AI models, you need to be able to explain why your algorithm made a certain decision. If you cannot, you fall afoul of GDPR restrictions.
“Furthermore, we anticipate additional regulations from payment providers such as Mastercard and Visa in 2024,” he adds. “While Visa’s forthcoming regulations are expected to align closely with Mastercard’s guidelines, merchants must remain vigilant and adaptable in response to regulatory changes. Staying informed through regular updates from payment providers is crucial.”
Beardsley says that Segpay will continue to monitor AI developments, and particularly how card brand regulations related to AI develop in the high-risk adult market.
“Card brands have already created policies around deepfakes and any sort of user-generated content that falls under Mastercard regulation AN5916, the Revised Standards for New Specialty Merchant Registration Requirements for Adult Content Merchants, which went into effect back in October of 2021,” she offers. “Visa then followed with the Visa Integrity Risk Program (VIRP) in April 2023. We also are paying close attention to the U.S. government. To date, there has been no federal legislation on AI. However, President Biden signed an executive order on AI outlining guiding principles to drive U.S. AI policy. Those include safety, innovation, protecting workers, protecting consumer rights, equality and investing in professionals.”
Due to the current lack of more specific federal legislation, state and local governments are tackling AI within their own legislative and regulatory framework, while outside the U.S., governments are implementing their own safety standards.
“The EU AI Act is intended to ensure the safety of AI systems on the EU market, and to provide legal certainty for investment and innovation in the AI field,” Beardsley explains. “It’s also designed to minimize associated risks to consumers as well as compliance costs for providers. The U.K. is also focusing on developing guidelines for AI.
“Currently, there is a variety of legislation and regulations related to AI in development, like data and consumer protection, product safety, equality law, financial services and medical device regulation,” she says. “At this point, there is no overarching framework that governs AI’s use.”
According to Beardsley, machine learning is already part of Segpay’s risk management toolkit, and the company aims to utilize AI to beef up that toolkit and become even more efficient.
“If AI is all the rage, we will need to up our game too,” she says. “Questions we ask ourselves are: Can we add AI to better support consumer requests and make our responses more efficient? Can we use AI to make our customer service even more personalized? For example, if we see a consumer repeatedly, should we have the ability to present different offers to the consumer when they opt to cancel?”
Stronger AI, Beardsley points out, will be able to analyze patterns in transactional data and help Segpay flag suspicious transactions in real time for further investigation.
“AI can help us analyze data and route transactions to optimize approvals for our merchants,” she elaborates. “It can also help us identify which banks outperform others in certain countries — and bolster network security, ultimately helping to keep out bad actors and stop potential bot attacks and distributed denial of service (DDoS) attacks.”
The Pandora’s Box of AI Legal Considerations
There are numerous legal considerations for both talent and business owners to consider when venturing into the murky waters of AI. For performers and creators, the goal is to protect their own rights and IP.
“The rapid evolution of synthetic media has allowed computers to generate images or video realistically depicting content creators engaging in acts they would never perform, or saying things they would never say,” observes Lawrence G. Walters of Walters Law Group. “All that is required is sufficient base content containing the creator’s image, likeness and voice, from which AI technology can learn and create. The possibilities for new synthetic media then become endless. Creators must therefore consider how to protect their brand, intellectual property rights and reputation from being negatively impacted by future AI creations depicting them.
“This starts with the model release,” Walters says. “Creators should pay careful attention to the scope of rights they are releasing in content production documents. Allowing the base images and audio files created during a shoot to be used for any purpose, in any media — existing now or to be created in the future — puts creators at risk. Given what we now know about the capabilities of AI technology, it is fair and reasonable for creators to negotiate limits on the use of their performance for future generative AI content.”
While the scope of copyright protection for synthetic media is unsettled in the courts and a hot topic of debate in IP law circles, Walters nevertheless advises creators to register all works they own with the U.S. Copyright Office within three months of publication. This, he says, will help them lock up maximum protection for their content and position themselves to provide potential legal arguments if necessary.
He also highlights “name, image and likeness” (NIL) rights as an effective tool for combating abusive AI content.
“Creators have the right to protect their face, appearance and voice from unauthorized commercial exploitation under a variety of state laws,” Walters notes. “Creators should educate themselves, or seek legal counsel, on how best to enforce NIL rights in the event unauthorized synthetic media is detected.”
Corey Silverstein of Silverstein Legal agrees, and urges all talent to treat their name and likeness as their most valuable asset.
“In the thick of content creation, consumer interactions, social media engagement and the myriad other tasks that performers face every day, it can be easy to forget or neglect protecting one’s name, likeness and intellectual property,” says Silverstein. “Mass adoption of AI has compounded this issue. The landscape of content creation has completely changed. For creators, understanding what rights they are maintaining versus providing to a third party has never been more complicated.”
In Silverstein’s estimation, conventional model release forms are now outdated. He sees this development leading to disastrous consequences, including complicated and expensive lawsuits.
“If you are a creator, the issue of AI usage must be discussed prior to the finalization of any formal agreement,” Silverstein emphasizes. “For example, if you are a performer and agreeable to a producer utilizing your name/likeness for the creation of AI-generated content, that needs to be carefully documented — and of course, the value of those added services should be carefully considered. Conversely, if you do not want your name and likeness used in any type of AI creation, or if you want to place specific limitations on such use, that also needs to be specifically documented.
“For talent utilizing the services of agencies, I urge extreme caution,” he warns. “Agencies utilize numerous variations on agreements that give them broad rights, including in relation to AI. I cannot emphasize enough the importance of utilizing industry legal professionals who can help talent understand potential contractual pitfalls.”
On the business side, paysite owners and user-generated content platform operators face a wide variety of potential legal challenges related to AI.
For instance, Walters notes that the scope of legal protection a platform enjoys in the U.S., in the form of Section 230 immunity, may well come down to what role the platform plays in the publication of AI content. If the platform does not produce AI content, but simply allows users to upload it, legal protections from most civil claims are at their strongest. He emphasizes, however, that website and platform operators should still make sure they have obtained the necessary legal rights to publish AI content, including any copyrights or NIL rights, from recognizable models depicted in such materials.
“If a platform is providing AI services that allow users to create synthetic media, the legal protections are less settled,” Walters explains. “Operators should prohibit users from creating materials that depict specific individuals without their permission. This is particularly important when it comes to celebrity content. The descriptions, tags or labels that users can utilize to create synthetic media on the platform should be carefully monitored to prevent users from generating unlawful materials or content that is deemed ‘brand-damaging’ by payment processors.”
Though copyright issues associated with AI content are muddy, Walters adds, platform operators should still stake out the legal protection offered by DMCA safe harbor in connection with user-submitted AI content.
“This requires filing a designation of DMCA agent with the U.S. Copyright Office, publishing a legally compliant notice and takedown policy, and adopting and reasonably implementing a repeat infringer policy,” he elaborates. “It is better to have DMCA safe harbor and not need it, than to need it and not have it.”
Walters also urges website operators to be highly attentive to FTC rules and payment processor policies regarding the proper identification and labeling of AI content, to ensure any promotion or sale is not deceptive — especially given the FTC’s broad oversight authority in connection with online promotional activities, including actions by affiliates.
“Adopting proper policies regarding influencer or affiliate conduct, and policing their activities, is essential to avoid burdensome investigations and expensive fines levied by the FTC in the event of suspected violations,” Walters says. “We anticipate an influx of investigations and civil claims regarding deceptive use of AI technology, so operators should get ahead of the curve by implementing proactive policies to mitigate risk.”
Silverstein notes that the all-powerful card brands are well aware of the explosion in AI content. Echoing the concerns of payment processors, he emphasizes that use of nonconsensual AI, like deepfakes, is unacceptable and a one-way ticket to losing the ability to process transactions.
“Consent is everything,” he affirms. “The use or creation of AI-generated content by adult businesses is legal — so long as properly documented consent has been established. Clear, unambiguous and specific language detailing how a business intends to create and use AI-generated content is essential in agreements with talent. Vague statements about how a performer’s name and likeness can be used by ‘all technological means in existence’ are not good enough.
“From an age verification standpoint, proper record-keeping is still a requirement if a business creates AI-generated content based on an actual human being,” Silverstein explains. “There is no age verification exception for AI-generated content.
“It is also essential for businesses to disclose to consumers whenever they are interacting with or purchasing content featuring AI-based models,” he continues. “Make no mistake: The FTC will be very active in claims for fraudulent and deceptive trade practices against businesses that do not make such disclosures.”
Platforms that do allow user submissions need to understand the risks associated with accepting AI-generated content, and update all TOS and similar agreements to properly address risks and liabilities, Silverstein says. He points out that many banking intermediaries are already prohibiting user-generated platforms from accepting AI submissions, so businesses need to have conversations with their credit card processors and banks about their policies related to AI.
“Just like talent, adult business operators utilizing AI-generated content should consult their attorneys to ensure that they fully understand their rights and responsibilities when it comes to AI usage,” Silverstein concludes.
AI Opportunities Abound for Innovators
Despite the many legal and payment issues it raises, plenty of adult industry folks are eager to tap into AI. The reason is simple: numerous and promising business opportunities.
“When I talk with people about AI-generated content, they immediately go to the most obvious place: completely AI-concocted influencers and content,” says Brian Shuster of Utherverse and Red Light Center. “To be sure, this kind of content does appeal to some people — but this is not where the industry will really feel the impact.
“By and large, consumers will always prefer content made by real humans,” he states. “It is the core of arousal. Knowing that an AI model is not real means that pleasure or pain or arousal or any possible reaction is simply a programmed response. It really detracts from the experience.”
What AI will provide, Shuster says, is a whole new set of tools to enhance content made by real people.
“Camera angles that were missed or impossible to obtain could be added to scenes in post,” he notes. “Lighting will be enhanced. In short, videos and stills will become hotter and more appealing with generative AI. AI can remove or alter the performers themselves. Blemishes can easily be removed, makeup added and new ‘mods’ of past scenes created where girl-next-door becomes goth, for example. Full face and body swaps could even open entirely new methods for customers to personalize their experiences. Imagine scenes in which one person’s face is swapped out for another’s — someone consenting, like a wife who would love her husband to watch her face on his favorite porn star.”
The most compelling and wide-ranging impact of AI, as Shuster sees it, will be the hybridization of real people and avatar models, which he likens to a real-life version of the “Star Trek” holodeck.
Bob Kelland of CloneTwin.ai does not foresee AI-generated images and video content meaningfully impacting the adult space for a few years yet — but when it does, he says, it will be due to the interactivity potential of AI-driven machine learning.
“If you are an adult creator, then AI is good news and bad news,” Kelland says. “First the bad news: Replicating a performer’s content using AI is becoming increasingly easy to do. Almost anyone can create hyperrealistic AI content that includes your voice and persona. Many of the creator agencies and platforms are looking to replace real creators with synthetics; it’s already happening.”
Kelland notes that such options are especially attractive to B2B outfits that complain about creators not having enough time and resources to do customs or chat engagingly with users.
“The good news: If you are a creator who is prepared to embrace AI, you can take control of your intellectual property and create your own AI persona,” Kelland continues. “It is now possible to have your own super-realistic AI replicating your persona, voice and content. Creators who get in early and create a hybrid version of themselves will have an advantage and do well in the AI world.”
Kelland thinks dating site and fan site operators are most at risk of disruption, but may also have the most to gain if, as he expects, AI outperforms current methods of chatting with fans and augmented real-time AI-generated content becomes a trend. For creators, he believes, analytics will also evolve to a point where AI can predict the ideal content to be released on any given day.
Then there is Steve Lightspeed of Porn.ai, who believes the future of adult is not about looking at what others have created, but sitting in the director’s chair and creating it for yourself.
“I don’t think it’s an understatement to say that everything about the adult industry is about to change,” Lightspeed says. “Many companies, including mine, are investing heavily in AI development because porn consumers have already shown a voracious appetite for AI-generated content. And it is only getting started! Within the next two to three years, we will see AI images and video continue to improve, becoming indistinguishable from reality. Ubiquitous and easy-to-use AI technology will turn everyone into a content creator, potentially wiping out the scarcity that supports the creators market today.”
Lightspeed has no doubt that AI will capture a significant share of the porn-viewing audience, and lists as evidence the technology's capacity for delivering high quality, endless variety, unlimited creativity, complete personalization, total control and 24/7 availability — benefits that are simply unbeatable.
“This doesn’t mean human performers are at risk of becoming obsolete,” he adds. “In my opinion, AI can never replicate the creativity and passion that so many performers possess.”
The AI user experience Lightspeed envisions, however, offers something else entirely. Instead of hunting through social media, tube sites, cam networks and search engines for content, he predicts, users will just manifest whatever they desire. He notes that users on his site already spend hours every day customizing and fine-tuning their AI creations.
“They put a very personal piece of themselves into their ‘AI girlfriends,’” he explains. “It may be difficult for the average OnlyFans model or cam performer to compete with that. Many users feel awkward sharing their most intimate moments with a stranger on the internet; they report a more comfortable experience with AI content.
“They already know that most performers don’t use their real names or real backstories, and that many are outsourcing fan mail and chatting to foreign chat factories,” Lightspeed adds. “If you’re not connecting with a real person anyway, why not just use AI?”
He believes performers will embrace AI by creating their own avatars and chatbots — but that companies may ultimately opt for entirely synthetic models rather than split profits with human performers.
Andra Chirnogeanu of Stripchat and My.Club offers an outlook likely to be more appealing to creators and performers. Like Lightspeed, she also foresees a merging of AI-generated content with human creativity, but one that will help creators deepen and broaden their work.
“We are already starting to see stories fully realized with AI-generated images, and videos enhanced with AI-crafted soundtracks,” Chirnogeanu says. “This collaboration is more than just layering on extra details; it’s a complete overhaul of the creative process, providing a fuller, more engaging storytelling experience.
“The introduction of AI in video generation, crafting visuals that mix digital with reality, brings limitless potential for storytelling,” she declares. “Creators can now design engrossing narratives that carry audiences to places beyond their wildest dreams.”
AI’s ability to customize content to an individual level, make changes based on real-time interactions and blend various types of media, Chirnogeanu believes, will transform the adult industry and the creator economy.
“The combination of human creativity with AI’s technical prowess is leading us to a future where the pursuit of ideas, themes and storytelling methods knows no limits,” she opines. “Look at the stats from platforms at the forefront of AI-driven content. According to SimilarWeb, JanitorAI.com draws 28.5 million monthly visitors, Candy.ai brings in 8 million, and SpicyChat.ai hooks 17 million, with users engaged for over 20 minutes per visit. This is on par with watching a compelling Netflix show or a playlist from your favorite YouTube creators, marking not just brief encounters but deep, engaging interactions.”
What other kinds of AI-generated content will make the biggest impact? Shuster sees AI bringing the best of the real world together with the best of AI, enabling fans to interact with avatars of performers in a personalized manner.
“Imagine an AI-powered avatar designed to look just like a particular performer,” he posits. “The real performer works with the AI company to train the AI avatar to move, talk and behave as that performer. Meanwhile, another company sells custom molds of top performers’ anatomy for their fans to use as masturbation toys.
“At that point, there is really a connection between the AI and the performer,” Shuster says. “It looks and acts like them, and the mold is from them — the fantasy connection is so complete that customers will no longer feel like they are just playing with a computer. Now they are watching, feeling and playing with a close replica of the actual person!
“This type of AI will give everyone a new way to create and profit,” Shuster concludes. “Models, performers and toy manufacturers will see a boom in their revenue from this application of AI.”
For Kelland, the biggest impact is likely to come from multimodal AI, which can recognize and generate text, images, audio and video in real time.
“Anyone can get content for free, so while AI video will be big, it is not a game changer in itself,” Kelland says. “The key is relating to the user so they feel engaged — and this is where multimodal AI will come of age sooner than you think. It behaves just like a real human. It can talk, text and share content, just like you would do with another person.”
Highlighting how fast AI technology is accelerating, Lightspeed cites a recent major breakthrough in text-to-video generation with OpenAI’s Sora, AliBaba announcing a new deepfake lip-syncing AI model, and Eleven Labs now making it possible to create voices that sound completely real. He says he is focused on building a community of adult AI enthusiasts who will collectively push the technology in new directions.
“We may not be able to make AI adult videos yet, but we intend to be very well-positioned, with millions of fans, when the technology inevitably arrives,” Lightspeed says. “We are currently perfecting the user experience and working through the technological challenges of making hyperrealistic AI imagery a reality. Interactive video with customizable synthetic performers is the holy grail of AI content. This is the sole mission of Porn.AI, and we have received backing from well-funded venture capitalists to make it happen by the end of 2025.”
By all accounts, AI is not a fad ripping through the headlines — NFTs! The metaverse! Crypto! — only to fizzle into a nostalgic curiosity. It is not a premature debut, all hype and no substance. AI is poised to transform every corner of adult. To those prepared to tap into it will go the spoils. But that preparation will require careful consideration of the many factors outlined above.