In the immediate aftermath of the Trump administration's gag orders on government employees disclosing taxpayer-funded research results, a series of high-profile "rogue" government agency accounts popped up on Twitter, purporting to be managed by civil servants who are unwilling to abide by the gag order.
At first blush, these accounts appear genuine in the main, though a few are apparent (and inconsequential) hoaxes. But it would be extremely easy to engineer a major hoax using these accounts, one that could have significant political fallout — from discrediting the opposition to Trump to generating panics or perpetuating scientific myths (the "rogue" EPA account could tweet evidence that climate change had been discredited, for example).
There's also the problem of remaining anonymous: if these account are run by civil servants who face retaliation in the workplace or even civil and criminal charges, then they run a significant risk every time they tweet. Not only might the Trump administration seek disclosure of identifying information through court orders, they might also deploy American intelligence agencies to covertly uncover this information through state-sponsored hacking. Less drastic but equally high-risk for the tweeters is the possibility that their adversaries will use stylometry to uncover their identities via the idiosyncratic elements of their prose style.
Then there's the matter of operational security: in 2012, the CBC de-anonymized a staffer in Canada's Liberal party who had been embarassing then-security-minister Vic Toews through the "Vikileaks" Twitter identity, which tweeted salacious details of Towes's divorce proceedings. The CBC sent the tweeter a private message advising him that they had published an article of interest to him, with a private, unpublished URL; they then watched their webserver logs for that page to see what IP address he visited from. That led to his unmasking.
The "rogue" administration accounts are operating in a much higher-profile/higher-risk environment that the Vikileaks account, and maintaining operational security, even for the very careful, is a hard row to hoe. Given that the tweeters' expertise is in other areas (for example, being a park ranger) it's unlikely they'll be able to maintain perfect security and even a single slipup could cost them their jobs (and possibly their freedom).
The LA Times' Matt Pearce has pointed to specific tweets posted by @AltNatParkSer — the most prominent of the renegade accounts — that could be seen as out of character for a US park ranger. British English is used in the biography, for one, while the account also replied to a Scottish political commentator in 2015, asking for his opinion on the UK general election result. Not that it matters much after the account announced this morning that it would be handing over the reins to activists.
None of this is evidence that the accounts in question are not run by US government officials, but nor is it confirmation that they were indeed run by scientists keen to spill secrets.
@AltNatlParkSer had said that it was based in Mt. Rainer National Park in Washington State, but wouldn't identify its authors because it didn't feel "safe" and it wasn't "in the public interest." Certainly the former point is a very valid concern, as going against government orders could easily result in loss of livelihood. But confirming that the most influential of these accounts are indeed the work of (rightfully) upset scientists arguably is in the public interest, because at the moment, there's no guarantee these supposedly renegade Twitter accounts are anything other than parody accounts offering wish fulfillment for progressives.
On the internet, nobody knows if you're a National Park
[Rich McCormick/The Verge]