Two Immigration and Customs Enforcement employees are under scrutiny for social media posts they approved or shared criticizing Democratic freshman Rep. Ilhan Omar after her comments about Israel.

Omar, a Muslim of Somali descent representing Minneapolis, sparked controversy in recent weeks for statements about Israel and the pro-Israel lobbying group American Israel Public Affairs Committee. At issue is whether her comments reflect anti-Semitism—which Omar refutes—but have nonetheless been widely debated, including within her own party.

The controversy centers on Omar’s perception of AIPAC’s influence in U.S. politics.

While the Democratic-controlled House reconsidered offering a resolution perceived to be an indirect condemnation of Omar, it did pass one specifically condemning anti-Semitism, anti-Muslim discrimination, and bigotry.

Now, Buzzfeed reports that at least two ICE officials have shared or approved posts on social media in recent days making erroneous allegations about Omar.

A senior ICE agent shared a post on LinkedIn that read, “This woman does not belong in Congress. She is a Trojan horse thanks to the refugee outbreak in Minnesota under Obama! She has already stated many dangerous things about our allies and her ideology. She must be removed!”

The post included a misleading video suggesting Omar had admitted to “taking terrorism classes.” The woman who posted it continued, “And this is what represents America?” The post was then liked by another ICE agent, who is the second-in-command of Homeland Security Investigations in San Francisco.

But while such actions could be deemed an “indicator of bias,” according to Buzzfeed, there is no strict policy prohibiting ICE agents from writing whatever they want on social media. An ICE spokesperson told Buzzfeed that the agency has no social media policy, but claimed that anything posted to an employee’s personal account “should not be considered representative of agency policies or positions.”

The spokesperson added that they are currently “reviewing whether any of this particular content violated other agency policies.”