May was the month a federal judge told the White House to cut it out.
They have to comply with the Presidential Records Act. Passed in 1978. It says the president’s papers belong to the public. Not them.
Just a month before that. The Justice Department argued the law was unconstitutional. A weird twist. The American Historical Association sued. American Oversight sued. They warned that this legal loophole would let officials hide behind personal email and encrypted chats. No accountability. A gap in history that never gets filled.
Judge John D. Bates thinks the law is likely constitutional.
Good for now.
But the court battle is only the surface. The real problem is the medium. Government decisions don’t happen in leather folders anymore. They live in email chains. Chat apps. Cloud documents. These are born in proprietary systems designed to vanish after a product cycle. Keeping them is a technical nightmare.
Volume is insane. The National Archives added 463 terabytes of electronic records in 2024. Just that year.
Mike Quinn, who runs Preservica, says it best.
The world is creating digital records at a pacing no organization anticipated.
First hurdle? The record has to actually survive long enough for archivists to touch it.
Laws say preserve it. Tech can do it. Smarsh claims to capture data from over 100 channels. Yet we see Cabinet officials talking military strategy on Signal. We see UK PM Keir Starmer allegedly using disappearing WhatsApp messages.
It slips away. Easily.
Private archives aren’t safer. Thorsten Ries at the University of Texas knows this. He uses digital forensics to dig into the past. Even when politicians or artists donate their physical papers to universities. The digital twin often gets lost. Overlooked.
Getting data off a hard drive without messing up the metadata is a skill. Rare skill. Different software versions save different file fragments. Different storage media keep different automatic backups. Those fragments tell you how a document was thought. How it evolved. But pulling that out is painstaking.
“This kind of knowledge,” Ries says. “Is actually still very sparse.”
Try getting it off a Google Doc without the password. Two-factor authentication blocks you. It is a locked vault.
But even if you have the file. Can you read it?
Software changes. Formats rot. Quinn says digital content doesn’t age like paper. It becomes obsolete. Unreadable. You have to migrate files. Convert spreadsheets. Move CAD designs to new versions. If you do it wrong. You misrepresent the original.
We saw that happen with the Jeffrey Epstein emails released by the Justice Department. Marred by rendering errors. The tech broke the record.
And if the file survives. If you can open it. It is still hard to use.
Copyright sits next to secret medical bills. Personal rants sit in the same folder as policy drafts. Archives get cautious. They gatekeep.
Lise Jaillant at Loughborough University points out the irony. A file exists on the server. Accessible from anywhere. Yet archives often still demand you be physically present. Travel. Wait. Comb through unfamiliar systems on borrowed time.
Volume kills speed. Jason R. Baron notes that the Freedom of Information Act (FOIA) is slowing to a crawl because of “staggering volumes.” Agencies search by keyword. They redact sensitive info.
It takes years. Sometimes over a decade. Just to get an answer.
AI might help. Maybe.
Baron explored this in 2025. Using machine learning to flag paragraphs that are part of an agency’s “deliberative process.” That shield. Software can spot Social Security numbers. OCR can pull text from old scans.
It finds what keyword search misses.
But there are gaps. Jaillant is leading a project on AI and government records. She says we lack good training data. Privacy laws mean researchers still rely on the Enron email corpus. It is decades old. Outdated.
And here is the thing. AI parsing does not replace reading.
It is still important for a human user… to understand the context.
You need a person to read the individual emails. To see what is happening.
All of this assumes the records survive.
The fight in Washington doubts that.
Archivists and their software are pushing hard. Trying to keep the records alive. Before today’s decisions get trapped in dead file types. Before they get erased from chat threads.
Before the public gets the chance to look.
