Report from Bloomberg
In Brief – UK Prime Minister Keir Starmer has announced plans to bring AI chatbots directly under the Online Safety Act (OSA) to close what he called a “legal loophole” in Britain’s online safety regime and ensure that they are designed to not produce harmful content. The illegal content rules of the OSA currently apply to user-to-user platforms such as social media but not AI chatbot services that respond directly to user queries. The move to regulate chatbots comes amid growing concern about abusive or harmful AI-generated content, such as the high-profile allegations that Grok, the AI chatbot of Elon Musk’s xAI, was used to create sexualized images of real people. Starmer called the material created by Grok “disgusting and shameful” and UK Ofcom, which regulates social media platforms under the OSA, opened an investigation of X for allowing the allegedly illegal images to circulate on their platform.
Context – The plan to pull AI chatbots under the OSA was presented by the UK Government as part of a broader effort to better protect teens online. The highlight is the prospect that the UK joins the queue of countries moving to follow Australia and institute a blanket age threshold for social media. The UK opened an expedited public consultation on a 16-year-old threshold in January. Social media age rules are reportedly being considered in at least 11 EU member states, including the French National Assembly recently passing legislation to set a 15-year-old limit and the Spanish government announcing plans for a 16-year-old threshold. Australia’s landmark age limit went into effect in December and one of the impacts was a rise in VPN usage. A similar spike in VPNs appeared in the UK last summer in response to a host of major online platforms implementing technical age verification as directed by Ofcom under the OSA. Officials in France, the UK, and several US states are calling for legal restrictions on VPNs to stymie their use in online regulation work-arounds.