Weak Keys Remain Widespread in Network Devices

Abstract

In 2012, two academic groups reported having computed the RSA private keys for 0.5% of HTTPS hosts on the internet, and traced the underlying issue to widespread random number generation failures on networked devices. The vulnerability was reported to dozens of vendors, several of whom responded with security advisories, and the Linux kernel was patched to fix a boottime entropy hole that contributed to the failures. In this paper, we measure the actions taken by vendors and end users over time in response to the original disclosure. We analyzed public internet-wide TLS scans performed between July 2010 and May 2016 and extracted 81 million distinct RSA keys. We then computed the pairwise common divisors for the entire set in order to factor over 313,000 keys vulnerable to the flaw, and fingerprinted implementations to study patching behavior over time across vendors. We find that many vendors appear to have never produced a patch, and observed little to no patching behavior by end users of affected devices. The number of vulnerable hosts increased in the years after notification and public disclosure, and several newly vulnerable implementations have appeared since 2012. Vendor notification, positive vendor responses, and even vendor-produced public security advisories appear to have little correlation with end-user security.

Machine Learning Cryptanalysis of a Quantum Random Number Generator

“…On the other hand, the guessing probability P g emerges naturally as the figure of merit by defining randomness as the unpredictability of a given distribution on a single use. This quantity tells us what is the best chance we have in predicting the outcome of a random (1), are followed by a LSTM layer, marked by (2), and two fully-connected layers, marked by (3). As a single input into RCNN, N (100) 13-bit integers are firstly encoded into one-hot vectors.…”

Network Hygiene, Incentives, and Regulation

“…Reporting vulnerabilities through Computer Emergency Response Team (CERT) organizations appeared to be of limited utility. Hastings et al found even more disappointing results in their study of the response to disclosure of an RSA private key compromise for 0.5% of HTTPS-enabled Internet hosts [20]. In a different study on remediation of hijacked websites, Li et al reported that browser interstitials and search engine warnings correlated with faster remediation compared to private notification via WHOIS contact alone [29].…”

Deep Learning-Based Security Verification for a Random Number Generator Using White Chaos

“…The output sequence of a DRNG is generated with a deterministic algorithm and a provided seed. Despite its good statistical characteristic, the DRNG is not suitable for information security applications, because the deterministic pattern of the DRNG may be identified by adversaries, which incurs malicious attacks and causes the destruction of security system, as in [ 2 , 3 , 4 , 5 ]. On the contrary, a NRNG produces the random sequence by using physical entropy sources, such as electrical noise [ 6 , 7 , 8 ], quantum fluctuations [ 9 , 10 , 11 , 12 ] and chaotic semiconductor lasers [ 13 , 14 , 15 ].…”