The SSL certificates issued by Israel based Certificate Authority StartSSL (https://www.startssl.com/) are blocked by Google Chrome and Mozilla Firefox since March 2017. Behind what could be just a technical issue, there is some disturbing facts:
First, the reason why Google and Mozilla have decided to progressively block StartSSL (and more importantly WoSign) is the issuance by WoSign, a chinese Certificate Autority, of multiple SSL certificate for Domains for which they didn’t received any mandate and didn’t validate the ownership of the domain by the requester. The first case to be reported to Google was GitHub, the famous Source Code repository. As WoSign had “secretely” bought StartSSL and integrated its infrastructure in its own, StartSSL has been “sentenced” to the similar distrust by most browser than its owning company.
As DNS CAA records are not used by browsers to check if the Certificate Authority of an SSL certificate for a domain is the correct one, it could have allowed someone to impersonate GitHub or at least to lure some users to a fake GitHub site (anyway, GitHub didn’t set his CAA record). Such behavior is unacceptable for any certificate issuer as trust is the cornerstone of the entire SSL certificate paradigm. Google and Mozilla’s reaction seems then proportionate. However, you can imagine the impact of such sentence. For any CA, being withdraw from the list of trusted certificates of the two main browsers is like a death penalty for the CA.
The second disturbing fact is that StartSSL failed (or decided not) to properly inform its customers. Worse, it continues to sell its Class 1 certificate despite the fact they are basically useless. That’s not the kind of commercial decision that will help restore the trust to the Israeli company, even if WoSign has defined a remediation plan aiming at giving more autonomy to StartSSL (see below).
Customers who had paid for the Enterprise Validation have lost their money and are now using blocking certificates. The only cheap and rapid solution to restore access to their website (and keeping the SSL/TLS active) is likely to use LetsEncrypt free certificates.
I don’t know what the future is but I wouldn’t recommend StartSSL to anyone anymore and I doubt any security aware person would. That’s not a good indicator for a bright future.
One famous saying attributed to Steve Jobs must be: “it doesn’t make sense to hire smart people and tell them what to do; we hire smart people so they can tell us what to do.”
It makes sense and security is no exception. How often do I see companies struggling to improve their level of security hiring external consultant while they have very talented and smart people capable of solving most of the issues… if you let them do it.
It might seem exaggerated but it is not so far from the reality. Your people may not have all the answers but they have likely solutions to a vast majority of your issues.
During lot of audit (or due diligence or GAP assessments), I interviewed managers and employees in order to get an idea of what works and what don’t in a company. Obviously, we check the incidents, the KPIs, the financial losses and all the possible indicators but its the discussion with the persons performing the jobs that give you the best insights. Rapidly, we can get a sense of where there is a bottleneck, a gap or an issue to fix. That’s normal, it is what we expect from external consultants. But what is often more surprising is that the same people are aware of the issues and have most of the time lot of ideas to fix them. It make sense as they are sometimes the persons suffering the most from these issues.
So, why are the issues still present? There is a lot of possibilities. One of the most common is the believe that the boss is always right (you know, rule #1). Hence, he likely know how to fix the problem, no reason to bother him with our stupid solutions. It creates blind spots. That’s probably why the space shuttle Columbia ended-up in ashes (see http://www.space.com/19476-space-shuttle-columbia-disaster-oversight.html).
Another possible reason is the difficulty of the people from the low level of the pyramid to talk the highest level’s lingo. Senior executives rarely want’s to have their hands dirty or to get involved in technical details or business processes considerations. I saw a few years ago a CIO meeting all the persons in its IT department (hundreds of people). Each meeting with a team gave him multiple hint on what was blocking or impacting the efficiency of his teams. And when you do, it’s easier to get the big picture and take the right decisions.
Another issue is the believe that the top management expect only green lights and positive outcome. “Failure is not an option” is a culture typically leading to failure. Also, sometimes, teams have opposed objectives, hence, they don’t work together to solves common issues but rather they fight each others or they continuously pass the hot potato. Not a good way to solve issues either.
A good and efficient security management, like any other corporate governance, requires an appropriate culture, fostering trust, empowerment, responsibility and so on. But these are more than words, they must be applied to be effective. bringing external consultants to fix internal issues is not always the best solution to improve your culture: it just send the message you don’t trust your team have the skills to do it.
You might want to try to express your expectations and discuss with everybody (or designated someone to do it) to figure out the best way to improve the situation. And if they need resources (what is likely the case) then maybe hire (external) people to reduce their current workload so they can start working on the changes.
Last tip: check your workforce’s skills… there’s sometimes people in your company who are doing work for which they are over-qualified and who could do jobs that could really provide you more added-value. Don’t look too far for your glasses, they might be on your nose.
When you’re working in the security industry, being paranoid is kind of natural (or is it the other way around?). So, when you see how easy people, processes and technologies can be hacked, you become rapidly suspicious of anything. We all know bad things can happen and most of the time we try to mitigate the risks (without even thinking too much about it). Business as usual, so to speak. However, while I have a good idea of the risks our future is bringing to us (what makes me even less worried about my business’ future), it seems that most people don’t imagine how much danger Internet will bring to them. So here are some clues.
The new buzzword that has a lot of attention in the media lately is probably IoT: The Internet of things. According to the media, it’s IoT who allowed hackers to put websites like Amazon and Netflix on their knee for a few hours on October 21st. But that’s a mistake. Although IoT has led to some specific new technologies like Bluetooth 4.1 or ZigBee to accommodate the low consumption and the low cost requirement necessary to embed technologies in nearly all objects, it is probably a mistake to see IoT like something new or something different. As Bruce Schneier said recently in front of the US congress, we should not see this has objects with computers in it (and an Internet connection) but rather see it as computer that do things. A Tesla is a computer with wheels (and when you see how Tesla manage its updates and is manufacturing process, it is closer to the Software industry than to the car industry way of working), a smartphone is a computer with a microphone and a 4G connection, a connected fridge is a computer with an extra cooling system, and so on.
Bottom line, these connected objects are all computers and we must treat them like it. So, like for all computers when it comes to managing security, we should think about patch management, access control, hardening, change management, release management, network segregation, encryption, key management, user awareness and training and all these processes and best practices. Unfortunately, the issue is that most connected object manufacturers didn’t spend enough time and money in designing secure objects, easily upgradable, with strong and secure communication protocols. Consequently, the future is now… and we are not ready for it.
But what is our future? Let’s get a glimpse at it. In the tenth episode of the second season of “Homeland”, Nicholas Brody help terrorists to kill a political figure by giving them his pacemaker serial number, allowing them to hack it and induce a heart attack.
In another TV show, “Blacklist”, a computer genius triggers remotely the airbag of a car while driving, causing the car to crash and the death of its driver.
Is this Science-Fiction? Unfortunately, not anymore! Exploits on “smart” cars become more and more frequent. More recently, a British and a Belgian researcher have devised a wireless wounding attack on pacemakers (1). While the latter exploit need specific and rather costly hardware (3 to 4.000€), we are just one step away of having a ZigBee or BT 4.2 interface. Do you wanna kill someone with your smartphone? Don’t worry, you won’t have to wait too long.
At the same time, as other device with less deadly capabilities are spreading over the world, they provide a potential army of unsecure devices that can be used for Distributed Deny of Service attacks, like it was seen recently, but, why not, to perform parallel tasking, helping to brute force passwords, crack cryptographic keys or hide communication sources by bouncing thousand of times on these little soldiers that we provide to these hackers. Nice isn’t it? We purchase the devices that will be used against us in the near future. To be honest, for most people, including for a lot of security specialist, it is not easy to make the difference between a secure IP camera and an insecure one, simply because we don’t have time to test everything and there is no useful and relevant certification for that. So think about the number of “computers” you have at home: Your internet router, you tablet, your PC or your Mac, your smartphones, your videosurveillance camera, your printer, your TV box, your Bluray player, your “smart” TV, your alarm, your new “connected” fridge, your smart thermostat, the PSP of your kids, the IP doorbell and so on… Think about it, in your home alone, you may have more than 10 little future soldiers for the next hacker’s army. Android, iOS or IP cameras, they nearly all have exploitable vulnerabilities.
So, we have an army and we have soon legion of potential targets for the new kind of attack: DoL attacks (Denial of Life). Imagine ransomware targetting your pacemaker, large scale attack on cars to cause traffic jams or worse, new hitmans (version 3.0) changing the medication of patients in hospital, overdosing people. Just watch any episode of “Person of Interest”, they were just a few inches away from the actual reality… and we are getting there.
It sounds crazy, isn’t it? As bruce Scheneier said, Internet is not that fun anymore. It’s not a game anymore. Things are getting serious and we should act accordingly. Not only at government level but also in industries and in the civilian world. We should ask our suppliers, our manufacturers to secure their devices, to make them safe AND easy to control.