Cryptologists recently cracked more than ten thousand certificates collected in the wild. These certificates are used to secure websites and email traffic. The group collected 6.4 million of such certificates. Analysis showed that these certificates have a lot more in common than they theoretically should have. That is, many certificates are not sufficiently random.

In their paper of 14 February Arjen Lenstra et al. analyzed a large collection of public keys. They discovered quite a few weaknesses in certificates, for example 22% of the certificates still use MD5 which is considered insecure and has even allowed rogue certificates to be generated (talk slides). Even more alarming is that they were able to crack more than ten thousand certificates.

The way they cracked the certificates is beautiful in its simplicity, requiring only high-school math and a modern computer. RSA certificates rely on the problem of prime-factorization: the certificate contains a very large number, the composite. The prime factors of this composite are unknown and form the secret key. The security of RSA is derived from the difficulty of finding the prime factors given the composite.

However, there is a well known method to find the largest common factor between two composites. The Euclidean algorithm, which is taught in high-school, can do this quickly. Lenstra et al. applied it to every pair of certificates to see if they have prime factors in common, a task thar a modern computer can do in a few hours. They found ten thousand prime factors common between a pair of certificates. This is a very alarming amount, since once one secret prime is revealed, the other one is easily derived and both certificates that are involved are completely broken.

Of course, the cryptologist designing RSA knew about the Euclidean algorithm and the possibility of this attack. That is why it is required to use numbers with thousands of bits and a very good source of random primes. As long as the primes are truly random the chance of two certificates sharing a prime would be negligible. The outcome of the paper has clearly demonstrated that plenty of certificates in practice are not sufficiently random.

The conclusion is a lesson we see time and time again: while the systems may be theoretically secure, it is important to implement them correctly. Even the smallest implementation flaw, in this case the random generator, will harm the security of the entire system.

Qubis heeft op zondag 5 februari in New York de Kairos 50 award ontvangen. De award wordt jaarlijks door de Kairos Society uitgereikt aan de vijftig meest innovatieve studentondermeningen ter wereld. De society heeft speciale aandacht voor ondermers die wereldwijde problemen oplossen. Qubis draagt met haar BMTool bij aan het verbeteren van cybersecurity. Het eventement vond onder andere plaats op de New York Stock Exchange, waar de jonge ondermers veel aandacht en advies ontvingen van succesvolle, gevestigde ondermers.

BMTool is een benchmarkomgeving voor cybersecurity, ontwikkeld in samenwerking met het CIO Platform Nederland, die met enthousiasme wordt gebruikt door haar leden, de academische en NVZ-ziekenhuizen en het Nederlandse hoger onderwijs. Naast Nederland is er ook internationaal veel interesse getoond.

Early morning of January 2nd, Qubis received an email from the Kairos Society.

Qubis gets to present itself at the 2012 Kairos Global Summit, held at the New York Stock Exchange.

We are proud of this achievement and would like to thank our clients for their trust in us. Without them, we would not have been able to achieve this.

Our new year could not have started with better news - we wish you all a successful, healthy and cheerful 2012!

Together with Iosif Androulidakis founder Yori Kamphuis submitted an abstract on Private Branch Exchange vulnerabilities for the 4th International Conference on Cyber Conflict. Today the NATO CCD COE invited us to proceed with our paper and submit the final paper before February 12th.

Qubis founder Remco Bloemen spoke at a hacking conference of the University of Groningen today. His talk was titled 'Veilige Informatiesystemen' ('Secure Information Systems'). His story began with the birth of computer science from early attempts at absolute certainty. These early pioneers started developing secure communication systems which led to the systems used today. However, if used incorrectly these systems can be exploited by hackers.

Topics discussed:

• Formal methods
  • David Hilbert’s program
  • Bertrand Russell’s Principia Mathematica
  • Kurt Gödel’s reaction
  • The birth of computer science
  • Alan Turing's codebreaking and Claude Shannon's Information Theory
• Symmetric keys
  • Electronic Codebook blockcipher mode
  • Cipher-block chaining mode
  • Initialization vectors (IVs)
  • Incorrect use of IVs in WEP and SSL
• Key derivation functions
  • Password entropy and salts
  • Feasibility of rainbow tables and hardware attacks
  • Defence against hardware attacks
  • Incorrect use of salts in WPA2
• Assymetric keys
  • Encryption / decryption
  • Sign / Verify
  • Certificates
  • Exploiting MD5 weaknesses in certificates
  • Central point of trust (and hence failure)
• Adi Shamir’s Secret Sharing Scheme
• Socialist Millionaire Protocol
• Of the record messaging
• Bitrate analysis of Skype

See also the homepage of 'Het hacking symposium' ('The hacking symposium') http://sympocie.svcover.nl/

Qubis has a new website. We hope you like it! A new item on our website is the weblog. We are going to publish various articles about Qubis and the IT-industry. Stay tuned for our first post!

On 14 November Qubis founder Yori Kamphuis received the message to have won the NATO Cyber Defence competition. Yori will travel to the NATO HQ in Brussels on November 28 to receive a Cyber Security briefing by the NATO Cyber Defence Section of the Emerging Challenges Division. Qubis congratulates Yori.