Twitter has agreed to give billionaire Elon Musk access to its “firehose” of raw data on the 500 million tweets posted each day, The Washington Post reports, citing a person familiar with Twitter’s thinking. Musk has demanded the data, alleging that the social media company he agreed to buy for $44 billion is underreporting the share of users that are automated “spam bots.” It is not clear if Musk will get full or partial access to the data “firehose,” The New York Times adds.

Musk has said on Twitter that his $44 billion purchase is “on hold” until he gets more data on spam and fake accounts, and his lawyers sent the company a letter Monday threatening to scuttle the deal unless Twitter hands over the data. The lawyers declined to comment on Wednesday’s reports Musk will get such access. Twitter pointed reporters to an earlier statement in which it said it “has and will continue to cooperatively share information with Mr. Musk to consummate the transaction in accordance with the terms of the merger agreement.”

The number of actual humans using Twitter, and thus the number of fake bots, is an important determinant of the company’s ad revenue, and it isn’t uncommon for a company to share proprietary data with a potential buyer. But when Musk “signed his initial deal to buy the company in April, he waived a right to look deeply at Twitter’s finances and internal workings,” the Post notes, adding that Twitter’s leaders, “along with some analysts and Silicon Valley insiders, say that Musk is using the data requests as a pretext to wiggle out of the deal or to negotiate a lower price.”