(Bloomberg) -- TikTok will “soon” grant Oracle Corp. full access to its source code, algorithm and content-moderation material as part of efforts to alleviate national security concerns about the app.

Oracle will also begin monitoring the controlled gateways where data comes in and out of the secure environment it set up on servers to host data from TikTok’s US users, according to a statement from the social media company Monday. 

The work is part of TikTok’s Project Texas, a plan to cordon off US users’ data and allow its technology to be reviewed by partners like Oracle to assess security risks. 

The social media company introduced the plan to alleviate criticism from policymakers who worry that its ownership by a Chinese tech company, ByteDance Ltd., opens the door to influence or data collection by the Chinese government. TikTok is currently facing federal- and state-level legislation that aims to ban or limit the app’s use in America on national security grounds.

“Many of the major components of Project Texas are already operational, and we will continue bringing more parts of the initiative online in the coming weeks and months,” the company said. Oracle started inspecting portions of TikTok’s source code in the so-called Dedicated Transparency Center, which went live in January. 

“Both TikTok and Oracle are continuing to work towards a solution with the US government,” TikTok said. “Our teams have been working together on all TikTok software that Oracle will ultimately inspect and continually monitor, while ensuring US users have an uninterrupted experience.”

Oracle didn’t immediately respond to a request for comment.

TikTok provided details on the arrangement after Bloomberg reported last week that Oracle had yet to embark on the most-involved elements of the program — intensive code reviews and app store updates — and won’t do so until the US government says the arrangement is sufficient. 

The Information added on Monday that Oracle employees only have limited access to TikTok’s source code and aren’t comprehensively reviewing its software, algorithm or content-moderation methods.

©2023 Bloomberg L.P.