Twitter won’t be treating President Donald Trump’s recent tweets telling congresswomen to “go back” to their supposed home countries as a violation of its hateful conduct policy, the company confirmed to The Verge. That means the tweets won’t trigger a flagging system Twitter announced last month, intended to limit the reach of banned content by public officials.

On Sunday, Trump attacked a group of congresswomen — implicitly the Reps. Ilhan Omar (D-MN), Rashida Tlaib (D-MI), Ayanna Pressley (D-MA), and Alexandria Ocasio-Cortez (D-NY) — by saying they came from “countries whose governments are a complete and total catastrophe” and telling them to “go back” to “the totally broken and crime infested places from which they came.” All four representatives are women of color, and three of them were born in the United States; as explained by Vox and The New York Times, the tweets invoke racist and xenophobic rhetoric that suggests non-white people and immigrants aren’t really Americans.

Gizmodo noted earlier today that Twitter’s hateful conduct policy bans “targeting individuals with repeated slurs, tropes or other content that intends to dehumanize, degrade or reinforce negative or harmful stereotypes about a protected category.” If Twitter had concluded that Trump violated this rule, it would theoretically have to hide the tweets behind a gray box that warned users about their content. That’s Twitter’s alternative to removing tweets that are a “matter of public interest” yet violate its rules.

It’s hard to tell where Twitter’s actual bar for “hateful conduct” sits, though, because content is moderated by large numbers of people who interpret the rules in different ways — as with most social media platforms. Also, Twitter in particular has a convoluted stance on racist content. You apparently can’t be a self-professed Nazi, but Twitter has said it’s talking with experts to figure out its more general rules around white supremacy.

Trump’s tweets have been widely condemned by political figures, including other Republicans. If Twitter intended to use its new feature to send a message about what’s acceptable on the site, it would have been an understandable moment to do so. As it stands, we still don’t know what would justify the gray box.