Why BBC has 'threatened' AI company led by Desi CEO

The BBC has threatened legal action against US AI search engine Perplexity in an effort to block the company from using its vast content archives to train AI models. The British national broadcaster alleges that Perplexity's AI model was trained on BBC content without permission. In a letter to Perplexity CEO Aravind Srinivas , seen by the Financial Times, the BBC stated it has evidence that the San Francisco-based startup's "default AI model" was "trained using BBC content."

The letter outlines a clear demand: the BBC could seek an injunction unless Perplexity immediately ceases scraping all BBC content, deletes any copies of the broadcaster’s material used for AI development, and provides "a proposal for financial compensation" for the alleged intellectual property infringement.

This move marks the first time the publicly funded British broadcaster has directly challenged an AI company on this issue.

Perplexity dismisses claims, says BBC wants to "to preserve Google’s illegal monopoly"
The Google rival, however, has rejected the BBC's allegations, calling the claims as "manipulative and opportunistic," asserting that the BBC has "a fundamental misunderstanding of technology, the internet and intellectual property law." Perplexity further contended that the BBC's actions were an attempt "to preserve Google’s illegal monopoly for its own self-interest."

It's important to note that Perplexity itself does not build or train foundational AI models. Instead, it provides an interface allowing users to choose between models developed by other major players, including OpenAI, Google and Anthropic. A source close to Perplexity indicated that its in-house model is derived from Meta's Llama and is refined to enhance accuracy and reduce "hallucinations," where AI models generate false information.