ChatGPT-like chatbots to fall into scope of UK law

AI chatbots are likely to come under scrutiny in the long-discussed online safety law.

Lord Stephen Parkinson, a junior parliamentary under-secretary in the Department for Culture, Media and Sport, was the one who confirmed plans to include AI-generated content within the scope of the proposed legislation.

This is because both search engines and social media platforms are in the process of integrating their service with software such as ChatGPT, exposing the potential risks of artificial intelligence tools.

The ultimate aim of the Online Safety Bill, currently under debate in the House of Lord, is “to make the UK the safest place in the world to be online” by holding tech executives accountable for breaking rules.

The bill must keep pace with new technologies

Content generated by artificial intelligence ‘bots’ falls within the scope of the bill where it interacts with user-generated content, such as on Twitter. bill to be described.’, said Lord Parkinson, The Telegraph reported (opens in new tab).

This probably comes as no surprise given the debate that arose after the first iterations of AI assistant tools being integrated into the services people use every day.

In particular, both the new OpenAI-based Bing and Google’s Bard have highlighted the vulnerability of this type of technology.

Despite the conversational nature of such language model systems, AI chatbots are not yet human enough to distinguish between what is true and false, what is good and harmful.

They could then expose users, especially children, to misinformation and potentially dangerous content – ​​exactly what the proposed bill aims to stop. GPT-style chatbots also seem to be prone to political bias.

Speaking of the decision, Lord Parkinson said: “The online safety law is designed to be technology neutral to make it future proof and to ensure the legislation keeps pace with emerging technologies.”

While the bill has long been criticized for its potential impact on freedom of expression online, its goal is to minimize exposure to illegal and harmful content online for both adults and children.

This includes potentially harmful material, such as material describing abuse, harassment, self-harm and eating disorders.

Lawmakers will also require social media and search engine platforms to be more legally transparent to their users about their content moderation practices.

Tech executives will face heavy fines and prison sentences if they fail to comply with the new regulations.

The review process was delayed amid the political turmoil that ended in late 2022. But now, after several amendments to the first draft, the Online Safety Act is currently under review in the House of Lords.