Elon Musk threatens to sue Microsoft over 'illegally' using Twitter data

Twitter CEO Elon Musk has threatened to take legal action against Microsoft over claims that the latter used data from Twitter without permission. The billionaire’s statement came in response to reports and tweets noting that Microsoft would drop Twitter from its advertising platform.

In a tweet on April 20, the billionaire said, "They trained illegally using Twitter data. Lawsuit time…

They trained illegally using Twitter data. Lawsuit time.

— Elon Musk (@elonmusk) April 19, 2023

Musk, who bought Twitter for $44 billion last year and has introduced some radical changes to the organisation since then, said that he is "charging businesses for access to its stream of data".

Earlier, Microsoft said in a notice that its advertising platform would "no longer support Twitter" from Tuesday 25 April. That means companies can no longer use Microsoft’s platform to manage their tweets or engagement. This also coincides with Twitter’s timeline to put its API (application programming interface) behind a paywall. 

Under Twitter’s new pricing arrangement, large companies like Microsoft could have to pay as much as $42,000 per month to gain access to Twitter’s API, according to a report in The Verge. The new pricing system has already led some smaller developers to abandon the platform, such as Tweetbot maker Tapbots, it said. 

Musk’s threat appears to be over Microsoft-backed AI firm OpenAI using Twitter data to train the large language model behind products like ChatGPT. Microsoft develops its own so-called large language models (LLMs) and sells access to OpenAI’s models. The company has invested significantly in OpenAI in recent months, and is building AI into tools like Bing, Edge, and Microsoft 365. 

That said, LLMs like GPT require terabytes of data for training, much of which is scraped from websites like Reddit, StackOverflow, and Twitter. Training data from social networks makes much better sense because it captures informal, back-and-forth conversations, believe experts. 

According to a report by New York Times, as the demand of these new AI models increase, these platforms are charging companies that are using them. For example, Reddit said earlier this week that it would charge companies for access to its programming interface used to feed the conversations among Redditors into AI training software.  

Meanwhile, in February, stock photo database Getty Images filed a case against Stable Diffusion alleging that the company copied its content to train its AI image generator. 

Earlier this week, Musk has announced plans to build his own large language model in one of his companies called TruthGPT, which will be "a maximum truth-seeking AI," And would challenge Microsoft-backed OpenAI and Alphabet’s (Google) Bard AI.