Britain’s national broadcaster, the BBC is sketching out plans to construct and train its own artificial intelligence model using its text archives, according to the corporation on Thursday, March 21.
A BBC executive had revealed to a parliamentary committee earlier this month that the organisation was looking at generative AI as a tool to boost its manufacturing process and assessing a likely unilateral or partnership arrangement on training a model.
An anonymous spokesperson had confirmed that the BBC was looking at developing a Large Language Model using text-based content. This spokesperson had also revealed that the output of any such model that they successfully build would only be available to the BBC.
The Financial Times had reported, while citing sources, that the broadcaster was in talks with technology firms over selling access to its archives, which could then be used to train the AI models.
In response to that, the BBC had said: “The BBC has no agreement with any organisation nor group to use its archive to train their large language models that power generative AI tools for public use.”
