Craig Wright is not Satoshi Nakamoto and the BBC should have known that

On 2nd May 2016 the BBC published a story titled “Craig Wright revealed as Bitcoin creator Satoshi Nakamoto”. Later the title was updated to the less confident “Australian Craig Wright claims to be Bitcoin creator”.

You can see the comparison between the two versions I posted on Github.

The first few paragraphs are especially telling. The following sentence:

His admission ends years of speculation about who came up with the original ideas underlying the digital cash system.

becomes later:

His admission follows years of speculation about who came up with the original ideas underlying the digital cash system.

(emphasis mine)

It’s great that a major news organisation is prepared to update the stories as its understanding evolves. You can see the full change history on News Sniffer. But the circumstances surrounding proving Wright’s identity as Nakamoto were so suspicious that the story in its first version should have never been published.

The BBC and other publications were given a scoop by Wright himself, but the validity of his claims were not confirmed by the journalists in person. His identity was verified by two experts instead. The experts were invited to a location chosen by Wright to check that he had access to a cryptographic key known to belong to Satoshi Nakamoto.

Nik Cubrilovic wrote a thorough article on all the reasons why Wright is not Nakamoto, and the part about the signature checking event is very thorough. The setup should have rang the alarm bells straight away. Nik’s post is worth a read if you’re interested in technical detail.

Wright generated a message requested by the expert, signed by a cryptographic key. The expert checked the validity of the message’s signature on a separate computer against the Bitcoin blockchain, to see if the signature matched that of the transactions known to belong to Nakamoto.

The laptops used to confirm cryptographic signatures were brought in by Wright’s associates (although on first inspection they appeared to be brand new and factory wrapped). The messages to be confirmed were passed around on a USB stick. The program used to confirm the cryptographic signatures of the messages was downloaded from the web using an unknown connection. In the type of the program used in the experiment the checks are done over the internet instead of checking against the local version of the blockchain. The experts were not allowed to keep the signed messages to independently verify them later.

Why was the experiment so suspicious? The hardware may have been tampered with, the network connection could have been set up by Wright or associates, and any downloads and checks made could have been done against a server set up to confirm the signatures.

The fact that Wright was unwilling to perform such a simple exercise under circumstances outside of his control should have alerted the journalists to the possibility that the confirmation was a staged trick.

Instead, the BBC took the confirmation at face value and published it as truth.

The problem with the first version of the story wasn’t the lack of knowledge about cryptography. The circumstances of the experiment should have been suspicious to anyone, even people without a good grasp of crypto (like me). The failing was about not having a basic understanding of information security. It’s important for journalists to not be so technologically naïve, if only to be able to protect their sources when their communications involve computers and the internet.