As today’s de-facto public square, the internet should be a place for discussion, not an echo-chamber.

That’s why new evidence that businesses are collaborating to suppress one side of a debate online is so concerning.

The House Judiciary Committee recently sent a letter to the World Federation of Advertisers’ (WFA) requesting more information about its member companies’ coordinated effort to demonetize and censor disfavored speech online. Ranking member Jim Jordan (R-OH) alleged in the letter that WFA’s subsidiary the Global Alliance for Responsible Media (GARM) colluded with several media and marketing companies to remove conservative speech from the internet.

As the letter reports, GARM built an alliance with internet giants, including more than 60 blue-chip advertisers and leading social media platforms such as Meta, YouTube, TikTok, and LinkedIn. GARM established uniform harmful content standards for its members that included vague and poorly defined terms like “hate speech,” “misinformation,” and “debated social issues.” And GARM members used these standards to censor or block advertising for content that didn’t fit their pre-determined narratives:

“Using these arbitrary standards, GARM works with its members to demonetize and eliminate disfavored content online. Indeed, GARM boasts that it “drove an agreement across advertisers, agencies and platforms to set a framework that limits advertising support” for disfavored content. GARM reinforces these collusive agreements not to advertise by “[d]riving mutual accountability, and independent verification and oversight.” Specifically, GARM’s aggregated reporting provides industry-wide data on how well social media platforms adhere to GARM’s standards. GARM’s efforts ultimately mean that even breaking news or op-ed coverage related to disfavored content may receive limited advertising support, and important content such as “[l]ive action footage/photos of military actions” and supposedly “[i]nsensitive, irresponsible and harmful treatment of debated social issues” may receive none at all. This collusive conduct reduces consumer choice and cuts off access to diverse coverage on matters of national interest.”

As Jordan later pointed out in an interview, conservative news publications or websites are the primary targets of these arbitrary standards. Organizations like GARM should not be deciding what speech is or is not acceptable. Their collusion across platforms has silenced valuable perspectives and caused conservative content creators to lose money. Furthermore, the House Judiciary Committee is currently examining whether GARM’s efforts violated a federal antitrust law against monopolies.

But GARM’s efforts are only one of many troubling threats to free speech online.

The letter to GARM comes on the heels of the House Judiciary Committee’s recent hearing on the Twitter Files, where journalists presented evidence that government actors worked with academic institutions and non-governmental agencies to wield the internet as a tool for censorship. And a recent series of investigative reports from The Washington Examiner showed that the U.S. State Department is funneling taxpayer dollars to a foreign activist group in the UK called the  “Global Disinformation Index” (GDI), which blacklists conservative news outlets in order to deny them advertising dollars.

A free society should not stand for censoring one side of important political and social conversations. Everyone loses when one side of a debate is forced out of the digital public square. The House Judiciary Committee’s oversight efforts are a necessary part of restoring accountability and respect for free speech.

In light of recent evidence, digital companies will need to be much more transparent about their practices in order to rebuild the public trust they’ve broken through their censorship. Tech companies should take a first step towards respecting the free speech rights of people on their platforms by responding to the survey portion of the Viewpoint Diversity Score Business Index.

These companies should publicly disclose any request or recommendation made by governments or non-governmental organizations to restrict content, along with the rationale for the request and the company’s response. They should also eliminate “speech codes” — policies that contain vague and imprecise terms that threaten free speech.