The company cuts off the CIA. But selling data to Vladimir Putin? That’s fine.
PHOTO: AGENCE FRANCE-PRESSE/GETTY IMAGES
34 COMMENTS
Silicon Valley’s hostility to U.S. intelligence and law enforcement reached a new low last week when Twitter rejected the Central Intelligence Agency as a customer for data based on its tweets—while continuing to serve an entity controlled by Vladimir Putin.
The Wall Street Journal broke the news that Twitter decided U.S. intelligence services could no longer buy services from Dataminr, which has a unique relationship with Twitter. Dataminr is the only company Twitter allows to have access to its full stream of hundreds of millions of daily tweets and sell the resulting intelligence to customers. Dataminr applies “big data” algorithms to identify unusual developments in real time. Customers who can profit from knowing about events instantly, such as hedge funds and news publishers, pay a hefty price for the alerts.
For the past two years, Dataminr provided its service to the CIA under a pilot program. The CIA and Dataminr then negotiated a contract to continue the service, but sources say Twitter CEO Jack Dorsey vetoed the contract at the last minute because he objects to the “optics” of continuing to help intelligence agencies. It’s unclear what happens to a small agreement Dataminr previously made with the Department of Homeland Security. With the new policy dictated by Twitter, Dataminr should drop the claim on its website that it includes “clients in the public sector, providing information first when there are lives at stake.”
MORE INFORMATION AGE
- Trump and the Obama Effect May 8, 2016
- How to Kill the Internet May 1, 2016
- Trump Models for Open Borders April 24, 2016
- Obama’s ‘Splinternet’ Legacy April 17, 2016
Among the customers still getting the Dataminr alerts is RT, the broadcaster created and funded by the Russian government. Vladimir Putin has said that the government runs RT to “try to break the Anglo-Saxon monopoly on the global information streams.” RT disclosed it is a Dataminr customer in its news account of Twitter barring the CIA. Agents of Russia’s Federal Security Service, formerly known as the KGB, have full access via RT to the alerts now being denied to the CIA.
A former top U.S. counterterrorism official last week explained to me the significance of the alerts Twitter has withdrawn. He said Dataminr informed intelligence agencies before the news broke about the Islamist attacks in San Bernardino, Paris and Brussels. “These real-time tactical warnings are especially useful in the case of multiple attacks when even if we only get a little bit of warning, it can make a big difference,” he said. “Dataminr alerted us ahead of the news in these cases by 15 minutes to half an hour, which meant we could make more effective responses to the attacks.”
The CIA owns 5% of Dataminr, through its In-Q-Tel venture-capital arm (Twitter also owns 5%). This is one of several investments the agency has in startups tracking open-source information such as tweets. A spokesman for the CIA reacted to the loss of the Twitter data by saying data from tweets were “critical in providing indications of pending plots” by Islamic State and al Qaeda.
The CIA’s No. 2 official, David Cohen, explained in a speech at Cornell in September that Islamic State’s “tweets and other social media messages publicizing their activities often produce information that, especially in the aggregate, provides real intelligence value.”
Twitter defended its decision by saying it doesn’t allow its data to be used for “surveillance purposes.” This is nonsense: Dataminr performs no surveillance, because there is nothing private about the tweets it mines for data. In any case, Dataminr uses the public information from Twitter to identify events, not to collect information about individual people.
Twitter’s argument raised eyebrows even among those usually sympathetic to arguments for privacy over security. “If Dataminr simply sorts public knowledge, then denying the intelligence community that information makes little sense,” explained Wired magazine. “If, on the other hand, Dataminr provides a level of insight that should only be accessible with a warrant, giving that same information to an unregulated hedge fund seems problematic.”
Unlike Twitter, Apple at least claimed to be protecting user privacy when it refused to help the FBI gain access to the iPhone used by one of the San Bernardino terrorists. But contrary to the dire warnings of Apple CEO Tim Cook about creating a “backdoor,” no privacy violations have been reported in the weeks since the FBI used third-party software to hack into the San Bernardino iPhone.
Silicon Valley seems to think consumers want it to pick fights with U.S. intelligence and law enforcement. Tech companies may be right that “optics” matter, but the optics now are that Twitter and Apple willfully interfere with America’s security even when no one’s privacy is at risk.
