As widely reported this week, the Trump administration has apparently sent memos to several federal agencies — including the EPA, HHS, the Interior Department, and the USDA — ordering them not to issue any press releases, publish new materials on their web sites, or post on their social media accounts. The memo to the EPA apparently added that “Incoming media requests will be carefully screened.”

As of this blog post, the memos are not yet public. But there is some disagreement among agency personnel who have seen the memos whether they represent anything new or unusual (or nefarious). For example, some say the memos are similar to ones sent by the Obama administration in 2009, and represent nothing more than a temporary “pause” or “reset” button as a new administration shifts policies from its predecessor. (Indeed, read together with the memorandum from Reince Priebus, “Regulatory Freeze Pending Review,” this has some purchase.) Others are calling the memo a “gag order” that portends an ominous shift toward censoring information from agencies that might conflict with President Trump’s controversial policy agenda, particularly on matters of health, science, and the environment.

Regardless of the memo’s novelty, many are suspicious that the Trump administration will repurpose this old tool to spread “alternative facts,” or at least suppress inconvenient facts. Given President Trump’s unconventional nominees to lead several agencies, a tool previously used by incoming administrations to centralize the vast public relations machine of the executive branch may take on new dimension. How long will the media blackout last? How selectively will it target agencies whose missions have been openly doubted by the President and/or his nominees? What legal basis does the blackout have, particularly if it contradicts statutory commands that agencies inform the public?

Already, the media blackout seems to be causing some intra-executive frictions. The Twitter feed for the Badlands National Park, part of the National Park Service and the Department of Interior, went rogue by “subtweeting” statements seemingly intended to defy the Trump Administration on climate change. Other arguably subversive tweets have been posted by the Department of Defense, one apparently in response to rumored executive orders limiting refugee entry into the United States.

These activities raise interesting wrinkles to arguments about the unitary executive, including who can and should control public statements by executive agencies. In the old days, agencies typed out press releases distributed to a list of reporters and mainstream media contacts. Today, hundreds of federal agencies have their own sprawling web sites, and make use of two dozen different social media platforms, including Twitter, Facebook, YouTube, and the like. (The site DigitalGov allows users to authenticate government social media accounts.) These platforms allow agencies to make public statements more quickly and casually than ever, and will obviously raise headaches for any administration trying to centralize and preclear communications.

But what governs agency communications? Both very much and very little. On one hand, the General Services Administration (GSA) runs DigitalGov, which maintains a handy “Checklist of Requirements for Federal Websites and Digital Services,” listing dozens of statutes, regulations, and guidelines (particularly from OMB) for agencies posting information online. The Obama administration also published a memorandum on scientific integrity in 2009 that directed science agencies to adopt policies to protect employee free speech rights and ensure public trust in government data. But the memorandum is not legally enforceable or very durable when a new administration changes priorities. Moreover, there are few rules that contemplate anything like the current media blackout. (Indeed, some agencies like the EPA and HHS have announced that they will continue to provide information to the public.)

Federal information policy is difficult because there are so many units communicating so much information to the public in so many different ways. Some communications are mandated specifically by statute. Some are justified under broad grants of authority to warn the public, protect public health, or prevent consumer harm. For example, could the Trump administration really object to the FDA communicating a salmonella outbreak to the public, or the CDC warning the public of further spread of the Zika virus? The vast majority of agencies could probably point to some statutory authority authorizing them to make information public.

What happens when presidential directives conflict with these statutory obligations? Could President Trump ignore statutory commands by Congress that agencies make certain information public? If this kind of “pause button” is normal, how long would it have to last to trigger real problems? How would we resolve the inter-branch separation of powers problem, or even intra-branch problems here? For example, when Congress passed a law directing the Department of Homeland Security’s Chief Privacy Officer to make reports “directly to the Congress, without any prior comment or amendment” by superiors, including the OMB (6 U.S.C. § 142(e)(1)), the Office of Legal Counsel objected on separation of powers grounds.

Obviously, the media blackout raises several important questions. The longer it lasts, the more aggressively it’s enforced, and the more selectively it targets certain agencies, the more these questions will demand answers.

Nathan Cortez is the Callejo Endowed Professor of Law and the Associate Dean for Research at the SMU Dedman School of Law. Follow Professor Cortez on Twitter: @nathancortez.