Understanding Twitter’s Security Enhancements
Last week Twitter announced a significant security enhancement – enablement of forward secrecy on twitter.com, api.twitter.com, and mobile.twitter.com domains. The prime objective of this change is to deprive hackers of an important exploitation tool.
Why was this necessary?
In the past 18 months, Twitter, akin to many other social network companies, has come under several cyber attacks by hacking community. Though not publicly acknowledged, some of them may have lead to data breaches as well.
Also in between a new adversary has emerged in the form of Govt. assisted programs like NSA. It is now widely believed that such programs have collected user data without consent or knowledge of companies.
Quite naturally these social network companies are concerned about the breach of users’ privacy and data. In addition to downtime or data loss the biggest risk from such attacks is loss of reputation and the trust of millions of subscribers. No business can afford to ignore concerns of such a large number of users and hence action was imminent.
How would forward secrecy help
It is a common practice that an adversary would capture the targets network traffic and mine it for crucial information. Plain text (i.e. not encrypted) traffic can be very easily skimmed for passwords and other PIIs. So Twitter and other companies decided to switch to HTTPS which meant that network traffic was travelling in an encrypted form.
But even this was not enough especially when the company used a limited set of secret keys to encrypt the traffic. Once the adversary obtained one or more such secret keys they could de-crypt the whole traffic captured earlier. To circumvent this risk Electronic Frontier Foundation (EFF) recommended the site to implement Forward Secrecy or also known as Perfect Forward Secrecy (PFS).
In forward secrecy the session keys used by servers are truly ephemeral. Twitter implemented it by enabling the EC Diffie-Hellman cipher suite. In this scheme the client and server manage to come up with a shared, random session key without ever sending the key across the network, even under encryption.
Due to this it becomes extremely difficult (though when it comes to security nothing is impossible) to de-crypt previously captured traffic and also does not help in guessing future encryption keys.
Is this all Twitter would have done towards improving security?
The recent announcement will help strengthen user confidence and stop them from exploring other options. The EFF had recommended implementation of additional security features like HTTP Strict Transport Security; secure cookies, encryption of data center links, STARTTLS, and certificate pinning and Twitter is doing good on that front.
Forward secrecy largely affects the traffic in transit, one of the three broad areas where security lapse can occur.
Another set of changes concern end users directly. HTTPS is something that end users might have noticed on their browser. Earlier in May this year twitter introduced improved login verification using two factor authentication. They also educate users on improved secure practices.
Finally Twitter must have taken measures for better in-house handling of data. Organization security policies and practices, cloud base backup and recovery plans and procedures or physical security come into ambit. This is something that only internal employees might have been informed about.
Does this all make Twitter future safe?
Not really. Security is always a dynamic system. Something that seems secure today may become vulnerable tomorrow. Not so long ago HTTPS alone was considered safe. These steps may be sufficient for now, but surely will get revisited in future.
Twitter thus joins other cloud companies like Facebook, Google and Dropbox in implementing these measures. With all leading cloud companies implementing it, soon this will become the default level of security.
By Manoj Tiwari