Can we trust web journalism?

I've spent the last couple of days meditating on the mistakes that web journalists make, and how those mistakes differ from mainstream media's errors. The reason for the meditation is a weirdly escalating cycle of misquotation that I experienced last week.

In general, I don't obsess about the mistakes that journalists make when I talk to them. If you get quoted a lot, you can expect to be misquoted a lot too, and it's best to let it go. Reporters are in a hurry; or their editors lack context; mistakes happen. Complaining feels a little whiny, and in any event, readers are likely to forget the story before a correction hits the wires.

But I was struck by the way this particular misquotation bounced around the web, acquiring authority by repetition without ever being verified, and I suspect it tells us something troubling about where the press is going, even for those of us who celebrate the breaking of mainstream media's narrative monopoly.

First, the background. I'm a skeptic about the Silicon Valley movement to increase the use of communications encryption that even the supplier can't undo. I think it's bad policy, and not particularly good business, for reasons I offered recently in a NYT op-ed:

That decision should not be left to Apple alone. And it won't be.

Companies do not want to give their employees the power to roam corporate networks in secrecy. And even if they did, their regulators wouldn't let them. If Apple wants to sell iPhones for business use, it will have to give companies a way to read their employees' business communications. Corporate IT departments won't welcome a technology that could help workers hide misdeeds from their employer.

And as a global company, Apple is subject to regulation and market pressure everywhere. If China doesn't like Apple's new policy, it can ban the iPhone or simply encourage China's mobile carriers to slow Apple's already weak sales there. Even democracies like India, and U.S. allies like the United Arab Emirates, have shown the determination and the clout to force changes in phone makers' security choices.

I repeated much the same view last week in Ireland, on stage with a Guardian editor, noting that Blackberry had run into real resistance in selling its end-to-end encrypted products in other markets. The Guardian wrote up the event in a somewhat sloppy story:

"Blackberry pioneered the same business model that Google and Apple are doing now - that has not ended well for Blackberry," said Baker.

He claimed that by encrypting user data Blackberry had limited its business in countries that demand oversight of communication data, such as India and the UAE and got a bad reception in China and Russia. "They restricted their own ability to sell. We have a tendency to think that once the cyberwar is won in the US that that is the end of it - but that is the easiest war to swim."

The sloppiness of the story shows in its still-uncorrected "easiest war to swim" error, but also in its framing. The Guardian's summary of my remarks begins with something I didn't say: "Baker said encrypting user data had been a bad business model for Blackberry, which has had to dramatically downsize its business and refocus on business customers." It's a plausible misunderstanding of my remarks, but it's wrong, as the Guardian could have easily found out during the many hours I spent that day with its reporter, James Ball.

What's striking is what happened next. The error went from plausible misunderstanding to outright mischaracterization. By the next day, several web outlets were using headlines like this one from ZDNet: "Former NSA's chief lawyer: BlackBerry's encryption efforts led to its demise." This is unequivocally wrong, both as a summary of my remarks and as a matter of fact, not least because Blackberry is far from demise. But it's also wrong more fundamentally; Blackberry's strongest market is selling to enterprises, many of whom are attracted to the product precisely because it offers very strong encryption that is controlled by the company and not by the individual user. It is, if anything, an illustration of why encryption is a lot more complicated than Silicon Valley's technolibertarian engineers seem to think.

That was not the end of the matter. There were soon a dozen or more web stories making similar claims, some from more or less respected outlets like Slate, TechSpot, the Daily Caller, the Register, and the Inquirer. Remarkably, many of them questioned the accuracy of blaming the "demise" of Blackberry on its strong encryption; in fact, they called that view everything from "strange" and "a bit of a stretch" to "absurd" and "laughable." What they didn't do was ask me whether I had actually made the claim that they considered so absurd and laughable. Not one of those web outlets called or wrote to ask for a followup quote or to confirm that I had made a statement they clearly thought no one in his right mind would make.

Why not? On reflection, I think it's because they liked the idea that someone on the other side of the crypto debate was saying dumb things about Blackberry and its encryption. The dumber the better, in fact. Call it the Twitterization of debate: Anyone we disagree with must first be caricatured as a dolt and then dismissed in 140 characters. Or call it the metastasization of the Huffington Post clickbait stylebook: "You won't believe this story showing how stupid/evil our opponents are!" Whatever, it's an understandable tactic if you're a partisan for a particular view. What's striking is how far that partisan style has infiltrated web outlets that to all outward appearances are engaged in, you know, journalism. (Indeed, it has much the same effect as actual journalism; within two days, reporters were asking Blackberry officials to respond to the still-unchecked quote.)

The "story that's too good to check" is part of newsroom lore, and an ever-present temptation for journalists. On the web, though, "too good to check" looks more and more like the norm, not the exception. And that's a problem for consumers of news. Sites like Slate and the Register present themselves as opinion journalism. We expect them to give us the facts along with the attitude. But increasingly it looks as though their facts are as open to question as their opinions.