With comments from- Larry Wong, Katten Muchin Rosenman
OpenAI plans to allow a wider range of content, including erotica, on its popular chatbot ChatGPT as part of its effort to “treat adult users like adults,” says CEO Sam Altman. In a post on X, Altman stated that upcoming versions of ChatGPT will behave in a more human-like way- “but only if you want it, not because we are usage maxxing.”
Lily Jamali and Liv McMahon writes for BBC that this move, reminiscent of Elon Musk’s xAI recently introducing sexually explicit chatbots to Grok, could help OpenAI attract more paying subscribers. However, it is likely to increase pressure on lawmakers to introduce stricter rules for chatbot companions.
In the UK, written erotica does not require age verification under the Online Safety Act, though pornographic images, including AI-generated ones, do. In the US, critics argue OpenAI’s move highlights the need for tighter federal and state regulations. Legal experts express concerns about ensuring children cannot access adult-only sections of ChatGPT, while surveys indicate that AI romantic interactions are becoming increasingly common among students.
California Governor Gavin Newsom recently vetoed legislation that would have blocked AI chatbot companions for children, citing the need for adolescents to learn to safely interact with AI. Meanwhile, the US Federal Trade Commission (FTC) has launched an inquiry into AI interactions with minors, and bipartisan Senate legislation would allow users to hold chatbot developers liable.
OpenAI, despite growing revenue, has yet to achieve profitability. Analysts note that the company is in a fierce battle for market share, aiming to sustain the rapid adoption of ChatGPT and achieve market dominance.
With this news, Larry Wong, Associate at law firm Katten Muchin Rosenman LLP, comments:
“Under the Online Safety Act, pornographic content consisting only of text (i.e. erotica content) does not qualify as pornographic content. However, the inclusion of any text-to-speech capabilities or synthetic audio content, such as those already available on ChatGPT, may result in erotica content being qualified as pornographic content for the purposes of the OSA, so long as the principal purpose of the work is for sexual arousal.
ChatGPT providing ‘qualifying’ pornographic content (i.e. providing content beyond just text) would result in it being qualified as a Part 5 provider pornographic content. Ofcom’s guidance makes clear that services making available generative AI tools to allow the creation of content through prompts or controls by a user will be treated as a Part 5 provider. As such, providers such as ChatGPT must have ‘highly effective’ age to prevent content harmful to children (such as pornography) from being accessed. Although details of the plans for ChatGPT’s age verification will follow, ‘highly effective’ age verification must be in place prior to the new erotica content service being made available to users.
The OSA takes a principled approach to what qualifies as ‘highly effective’ age verification, systems used by ChatGPT must be:
i) technically accurate;
ii) robust;
iii) reliable; and
iv) fair.
To avoid cutting off users from the rest of the service which does not contain content harmful to children, ChatGPT should only ‘age-gate’ its adult content services. This may be an opportunity for ChatGPT to develop an ‘adult section’ of the service that only age verified users can access. This creates opportunities to beyond erotica but also to a search engine service for adult content, which ChatGPT does not currently provide.
AI platforms have been under scrutiny for the content their models generated, particularly content that encourages or assists suicide. Whilst the OSA does protect users online, it is limited in scope and there is room for legislation specific to AI governance to complement what the OSA is trying to achieve and ensure that users are protected consistently online from AI systems.”
