OSA expansion to cover AI chatbots is a win for online safety

Today the Government has announced new proposals to reduce harm online, a welcome step that recognises the need to further strengthen, implement, and enforce the Online Safety Act.

We are pleased to see proposed measures to ensure AI chatbots are included in the OSA, AI chatbots are linked to a myriad of harms at both an individual and a societal level, including psychological harms, child sexual abuse material (CSAM), violence against women and girls (VAWG), disinformation and harms to democracy during elections. We look forward to seeing the detail of these proposals to ensure the full scale and breadth of harms are considered.

We stand with parents demanding swift action from the Government. We therefore welcome measures on the addictive design of online spaces and products, which go beyond the current focus on illegal content to tackle harmful platform design. The onus must be on tech platforms to make their platforms safe by design by strengthening obligations for companies to undertake and act on their risk assessments before they release products to market, as laid out in our 10 point plan to strengthen the OSA. We emphasise, however, that concerns about addictive design are as relevant to chatbots as they are for social media.

Significant failures of tech platforms to take users’ safety seriously, such as the recent tidal wave of deepfake abuse of women and children on X’s image-generating chatbot ‘Grok’, has confirmed what parents and the sector have been saying for a long time; the UK mustn't wait any longer to protect all users from online harms.

We appreciate the urgency of these issues and recognise the significance of the steps forward proposed today. We urge the Government to ensure that these vital developments are introduced through new primary legislation, rather than statutory instruments, to allow proper scrutiny, and to fulfil parliament’s ambition to make the UK the safest place to be online.