Britain’s national broadcaster, the BBC, is embarking on a journey into the realm of artificial intelligence (AI), considering the construction and training of its own AI model utilizing its vast text archives, the corporation disclosed on Thursday.
A BBC executive revealed to a parliamentary committee earlier this month that the organization is exploring the utilization of generative AI to bolster its production process, evaluating potential unilateral or partnership arrangements for model training.
Addressing a report from the Financial Times on Thursday suggesting the BBC’s intention to develop its own AI models, a spokesperson confirmed the corporation’s interest in crafting a Large Language Model using text-based content.
The spokesperson emphasized that any output from such a model would remain exclusively available to the BBC.
Additionally, the Financial Times, citing sources, reported that the broadcaster is engaged in discussions with technology firms regarding the potential sale of access to its archives for AI model training purposes.
In response, the BBC clarified, “The BBC has no agreement with any organization to use its archive to train their large language models that power generative AI tools for commercial use. We are examining various issues, such as potential bias within such models, and how the BBC, either through partnership or unilaterally, can address these concerns.”