The Snowden Revelations: When Mass Surveillance Stopped Being a Theory

In the first week of June 2013, in a hotel room in Kowloon, a 29-year-old American named Edward Snowden sat across from a documentary filmmaker and a pair of journalists and began handing over a cache of classified documents he had carried out of a National Security Agency facility in Hawaii. He had chosen Hong Kong because he thought it gave him a little time. He kept his laptop under a red hood when he typed passwords, because he was worried about cameras. Within days the first stories ran, and a claim that had lived for a decade in the margins — that the United States government was collecting the communications records of ordinary people at industrial scale — moved out of the margins and onto the front page of every newspaper on earth, carrying the government’s own slides as evidence.

The programmes that were real

The temptation, before June 2013, was to treat mass domestic surveillance as the fixation of cranks. After June 2013 that was no longer tenable, because the argument was no longer being made by cranks. It was being made by documents stamped TOP SECRET, published in the Guardian and the Washington Post, and confirmed — grudgingly, partially, but genuinely — by the officials whose programmes they described.

The first revelation set the tone. On 5 June the Guardian published a secret order from the Foreign Intelligence Surveillance Court compelling the telecommunications company Verizon to hand the NSA the “metadata” of essentially all its customers’ calls — the numbers dialled, the time and duration of every call — on an ongoing daily basis. Not the calls of named suspects. Everyone’s. The order was real, it bore the court’s seal, and its existence forced the admission that this was a standing programme, one that ran continuously, later confirmed to reach across the major carriers, authorised under Section 215 of the Patriot Act and reauthorised in secret for years.

The next day came PRISM. A set of NSA slides described a programme collecting content — emails, chats, stored files, video calls — with the compelled cooperation of the largest American technology companies: Microsoft, Google, Yahoo, Facebook, Apple and others, brought in one after another between 2007 and 2012. The companies pushed back on the framing, insisting they had not granted “direct access” to their servers, and the precise plumbing was argued over for months. But the core fact held: under Section 702 of the FISA Amendments Act, the NSA was reaching into the accounts held by the world’s dominant platforms to collect the communications of foreign targets — and, unavoidably, of the Americans and everyone else those targets talked to.

The disclosures kept coming for the better part of two years, through the reporting of Glenn Greenwald, Laura Poitras, Barton Gellman and others. There was upstream collection: the NSA tapping the fibre-optic cables that carry internet traffic, at the physical points where they enter the country. There was MUSCULAR, a joint operation with Britain’s GCHQ that secretly tapped the private links between Google’s and Yahoo’s own data centres — collecting from inside the companies’ networks, behind the encryption they showed the public. There was XKeyscore, an analytical system a Snowden document described as letting an analyst search vast stores of collected internet activity — browsing, searches, emails — with a query and a justification typed into a box. There was Britain’s Tempora, buffering days of traffic pulled straight from transatlantic cables. There was Boundless Informant, the NSA’s own tool for counting how much it collected, which recorded billions of records a month.

None of this was ambiguous, and this is the part any honest account must concede in full. The suspicion that intelligence agencies were vacuuming up ordinary communications had been, for years, treated as a mark of paranoia. The Snowden archive demonstrated that the paranoiacs had, in substance, been describing the architecture correctly. The agencies were doing it, at a scale that outstripped what most of the critics had dared allege, and they were doing it under legal authorities and court orders the public had never been allowed to see.

The lie under oath

What sharpened the revelation into a scandal was the deception that had shielded the collection, as much as the collection itself. In March 2013, three months before Snowden surfaced, the Director of National Intelligence, James Clapper, sat before the Senate Intelligence Committee. Senator Ron Wyden — who knew, from his classified briefings, what the answer was — asked him directly whether the NSA collected “any type of data at all on millions or hundreds of millions of Americans.” Clapper answered, “No, sir. Not wittingly.” The Verizon order published in June showed that answer to be false. Clapper later called it the “least untruthful” reply he felt he could give in an open setting, an explanation that satisfied almost no one and captured, in three words, the whole problem: the officials sworn to oversee the programmes had been misleading the public about their existence.

The Clapper exchange did more damage than any single slide, because it collapsed the defence that the programmes were secret only for security’s sake. A secret can be kept honestly; a false answer under questioning is a choice to deceive the very body charged with oversight. Wyden had given Clapper the question in advance and offered him the chance to amend the record afterwards, and the answer stood uncorrected until the Verizon order made it untenable. For a public trying to decide whether the intelligence establishment could be trusted to police itself, that single word — “wittingly” — did more to erode confidence than the existence of the programmes it was meant to conceal.

That is the documented spine of the affair, and it is damning enough on its own. Two independent reviews commissioned after the leaks — a presidential review panel and the Privacy and Civil Liberties Oversight Board — concluded that the bulk telephone metadata programme had produced little unique intelligence value and rested on a strained reading of the law. In 2015 a federal appeals court found that the Section 215 programme had never actually been authorised by the statute the government cited, and Congress, later that year, passed the USA Freedom Act to end the NSA’s bulk holding of the records. A theory had not merely been vindicated; it had been vindicated, reviewed, and legislated against, all within two years.

Where the record ends and the fear takes over

Here is the fork, and it is a delicate one, because the temptation after a vindication this total is to assume the documents endorse every fear anyone ever had. They do not, and the distinction is the whole discipline of reading them honestly.

The programmes the archive proved were programmes of collection, retention and search, governed — loosely, secretly, sometimes lawlessly, but governed — by statutes, court orders and internal rules that were themselves frequently broken. What the archive did not prove, and what the wilder retellings quietly smuggle in, is the fantasy of the seamless, omniscient state: an agency actively watching every citizen in real time, listening to every call as it happens, reading every mind, missing nothing. The reality the files describe is messier and more human. Collection outran the capacity to analyse; the agencies drowned in data they could not read. Internal audits, some of them also leaked, showed thousands of instances a year of analysts breaking the rules — querying data they should not have, including a category insiders reportedly nicknamed for the practice of spying on lovers. That is not the portrait of a flawless panopticon. It is the portrait of a vast, leaky, over-collecting bureaucracy that gathered far more than it could ever use and policed its own conduct badly.

The mythologised version also flattens a crucial distinction between capability and universal targeting. The NSA had the ability to reach an astonishing range of communications; the documents show it. It does not follow that it was individually monitoring you, or your neighbour, or any particular person absent a reason. The metadata programme collected records about nearly everyone precisely so that it could later query the small subset it claimed to care about — which is its own serious civil-liberties problem, the building of a haystack in order to keep the needles, but it is a different problem from the one the folklore imagines, where every straw is being read as it lands. Getting that difference right is not a defence of the programmes. It is the only way to argue about them accurately, and accuracy is the first thing the panopticon fantasy throws away.

The theory that had been true all along

There is a longer history here, and it is why the Snowden files landed on ground already prepared. Americans had been told once before, in detail, that their intelligence agencies had turned inward against them. The Church Committee of 1975 documented the NSA’s earlier bulk interception of telegrams under a programme called SHAMROCK, the FBI’s campaign of surveillance and sabotage against domestic dissidents under COINTELPRO, and the CIA’s mail-opening — a whole architecture of illegal domestic spying, exposed and supposedly reformed with the very FISA court that, decades later, would sign the Verizon order. The suspicion Snowden confirmed was not new paranoia. It was the memory of an old, proven betrayal, telling the public that the promise made in 1978 had quietly rotted.

That is what the “when it stopped being a theory” framing really points at. Mass surveillance had been a theory only in the sense that it lacked current documents. In the sense of precedent it lacked nothing; the precedent was overwhelming. What Snowden supplied was the missing decade of paper — the proof that the pattern had resumed, industrialised, and been hidden behind a court no citizen could petition and officials who would lie to the Senate to keep it hidden.

Snowden himself remains a contested figure, stranded in Moscow since 2013, called a whistleblower by some and a traitor by others, charged under the Espionage Act, unable to make the public-interest defence that would let a jury weigh whether the leaks did more good than harm. That argument will run for a long time. What is no longer arguable is the thing his documents settled. The claim that the government was collecting the communications of ordinary people at vast scale, in secret, on a strained reading of laws the public was not allowed to see, was not the delusion of the tinfoil-hat set. It was a description of the architecture, filed before the architects would admit the building existed. The lesson is not that every fear is warranted. It is the harder one: that a suspicion can be dismissed as paranoia for years and be, all the while, simply an accurate account of what was happening in a room no one was allowed to enter.