BBC to explore generic AI in journalism but block OpenAI web crawlers

The British Broadcasting Corporation (BBC) has said it will explore the use of generative artificial intelligence (AI) in journalism, but will not replace human talent or allow OpenAI to scrape its content.

The BBC is looking to explore how generic AI can help both journalists and news readers, Rhodri Talfan Davies, the BBC’s nation director, said in a blog post on Thursday. In the coming months, the news company plans to try out generative AI in areas such as news reporting and production, content discovery, maintaining archives, and creating personalized experiences.

However, the BBC also acknowledged the issues caused by new technology, which exploded in popularity this year due to chatbots like ChatGPT and text-to-image generators like MidJourney. These challenges include copyright problems, misinformation, potential bias, and legal or ethical risks.

For example, we do not believe that the current ‘scraping’ of BBC data without our permission to train general AI models is in the public interest and we want to agree a more structured and sustainable approach with technology companies. That is why we have taken steps to prevent web crawlers such as Open AI and Common Crawl from accessing BBC websites,” the BBC said in the post.

(For the day’s top technology news, subscribe to our tech newsletter Today Cash)

The news company promised to act in the public interest while adopting AI and said it would ensure transparency. It plans to use humans to publicly identify AI-powered content and monitor processes involving generative AI. The BBC also said that it would not replace human talent.

“We will always prioritize talent and creativity – no technology can imitate or replace human creativity. We will always prioritize and reward authentic, human storytelling by journalists, writers and broadcasters who are the best in their fields. We will work with them to explore how they can use generative AI to help push new boundaries,” the BBC said in the blog post.