I’m in a symposium at Churchill College on the Investigatory Powers Bill. It’s organised by John Naughton and I’ll be speaking later on equipment interference, a topic on which I wrote an expert report for the recent IP Tribunal case brought by Privacy International. Meanwhile I’ll try to liveblog the event in followups to this post.
Your browser contains a few hundred root certificates. Many of them were put there by governments; two (Verisign and Comodo) are there because so many merchants trust them that they’ve become ‘too big to fail’. This is a bit like where people buy the platform with the most software – a pattern of behaviour that let IBM and then Microsoft dominate our industry in turn. But this is not how trust should work; it leads to many failures, some of them invisible.
What’s missing is a mechanism where trust derives from users, rather than from vendors, merchants or states. After all, the power of a religion stems from the people who believe in it, not from the government. Entities with godlike powers that are foisted on us by others and can work silently against us are not gods, but demons. What can we do to exorcise them?
Do You Believe in Tinker Bell? The Social Externalities of Trust explores how we can crowdsource trust. Tor bridges help censorship victims access the Internet freely, and there are not enough of them. We want to motivate lots of people to provide them, and the best providers are simply those who help the most victims. So trust should flow from the support of the users, and it should be hard for powerful third parties to pervert. Perhaps a useful mascot is Tinker Bell, the fairy in Peter Pan, whose power waxes and wanes with the number of children who believe in her.
We are looking for three more people to join the Cambridge security group. Two job adverts, intended for postgrads or postdocs, are already out now. A third one, specifically aimed at a final year undergraduate or master student, strong on programming but with no significant work experience, is currently making its way through the HR pipeline and should appear soon. Please pass this on to anyone potentially interested.
- User experience (UX) designer
Research Associate or Assistant (with/without PhD)
Start date: ASAP
Details and link to application form: http://www.jobs.cam.ac.uk/job/9244/
- Senior software engineer / software engineer
Research Associate or Assistant (with/without PhD)
Start date: ASAP
Details and link to application form: http://www.jobs.cam.ac.uk/job/9245/
- Software engineer
Research assistant (having just completed a bachelor or master in CS/EE)
Start date: June 2016
Watch this space: the ad should go live within a week or so
This year, I’m on the PC for LASER 2016: the Oakland-attached workshop on Learning from Authoritative Security Experiment Results. The LASER 2016 CFP is now online, with a focus on methodologies for computer security experimentation, new experimental approaches, unexpected results or failed experiments, and, more generally, consideration of how to standardise scientific approaches to security research. Please consider submitting a paper — especially if you are pushing the boundaries on how we conduct experiments in the field of computer-security research!
The deadline is 29 January 2016. A limited number of student scholarships will be available to attend.
A very exciting Passwords 2015 is being hosted at the Computer Laboratory from 7 to 9 December. A unique conference that brings together the world’s top password hackers and academics. It is being liveblogged by the participants on Twitter with the hashtag #passwords15. A live feed is available on the Passwords 2015 page.
This afternoon at 4.30 I have been invited to give evidence in Parliament to the Joint Select Committee on the Investigatory Powers Bill.
This follows evidence I gave on the technical aspects of the bill to the Science and Technology Committee on November 10th; see video and documents. Of particular interest may be comments by my Cambridge colleague Richard Clayton; an analysis by my UCL colleague George Danezis; the ORG wiki; and finally the text of the bill itself.
While the USA has reacted to the Snowden revelations by restraining the NSA in various ways, the UK reaction appears to be the opposite. Do we really want to follow countries like China, Russia and Kazakhstan, and take the risk that we’ll tip countries like Brazil and India into following our lead? If the Internet fragments into national islands, that will not only do grave harm to the world economy, but make life a lot harder for GCHQ too.
We know more and more about the financial cost of cybercrime, but there has been very little work on its emotional cost. David Modic and I decided to investigate. We wanted to empirically test whether there are emotional repercussions to becoming a victim of fraud (Yes, there are). We wanted to compare emotional and financial impact across different categories of fraud and establish a ranking list (And we did). An interesting, although not surprising, finding was that in every tested category the victim’s perception of emotional impact outweighed the reported financial loss.
A victim may think that they will still be able to recover their money, if not their pride. That really depends on what type of fraud they facilitated. If it is auction fraud, then their chances of recovery are comparatively higher than in bank fraud – we found that 26% of our sample would attempt to recover funds lost in a fraudulent auction and approximately half of them were reimbursed (look at this presentation). There is considerable evidence that banks are not very likely to believe someone claiming to be a victim of, say, identity theft and by extension bank fraud. Thus, when someone ends up out of pocket, they will likely also go through a process of secondary victimisation where they will be told they broke some small-print rule like having the same pin for two of their bank cards or not using the bank’s approved anti-virus software, and are thus not eligible for any refund and it is all their own fault, really.
This paper complements and extends our earlier work on the costs of cybercrime, where we show that the broader economic costs to society of cybercrime – such as loss of confidence in online shopping and banking – also greatly exceed the amounts that cybercriminals actually manage to steal.
A lot of people are starting to ask about the security and privacy implications of the “Internet of Things”. Once there’s software in everything, what will go wrong? We’ve seen a botnet recruiting CCTV cameras, and a former Director of GCHQ recently told a parliamentary committee that it might be convenient if a suspect’s car could be infected with malware that would cause it to continually report its GPS position. (The new Investigatory Powers Bill will give the police and the spooks the power to hack any device they want.)
So here is the video of a talk I gave on The Internet of Bad Things to the Virus Bulletin conference. As the devices around us become smarter they will become less loyal, and it’s not just about malware (whether written by cops or by crooks). We can expect all sorts of novel business models, many of them exploitative, as well as some downright dishonesty: the recent Volkswagen scandal won’t be the last.
But dealing with pervasive malware in everything will demand new approaches. Our approach to the Internet of Bad Things includes our new Cambridge Cybercrime Centre, which will let us monitor bad things online at the kind of scale that will be required.
That’s the title of my PhD thesis, supervised by Markus Kuhn, which has become available recently as CL tech report 878:
In this thesis I provide a detailed presentation of template attacks, which are considered the most powerful kind of side-channel attacks, and I present several methods for implementing and evaluating this attack efficiently in different scenarios.
These contributions may allow evaluation labs to perform their evaluations faster, show that we can determine almost perfectly an 8-bit target value even when this value is manipulated by a single LOAD instruction (may be the best published results of this kind), and show how to cope with differences across devices, among others.
Some of the datasets used in my experiments along with MATLAB scripts for reproducing my results are available here: