Summary
• Cryptocurrencies
• Single track
• Interdisciplinary
• Keynotes:
– Adi Shamir
– David Chaum
Adi Shamir‘s 15-year predictions
• “RC4 and SHA-1 will fade out“
• “AES and SHA-2/3 will remain secure“
• “ECC may fall out of favor“
• “The Internet of Things will turn out to be a security disaster“
• “Cyber warfare will be the norm“
• “The situation will get much worse“
User Experiences withBitcoin Security & Privacy1
• Comprehensive user study with Bitcoin users(N=990, 8000 BTC)
• Topics
– Bitcoin management
– Anonymity, Security, Privacy
– Lost coins
– Usability
User Study Results
• 32% think that Bitcoin is fully anonymous
• 25% use Bitcoin over Tor
• Cumulative lost BTC: 660.6873
– HYIPS, risky investments
– Hardware failure, software failure,
– Major security breaches (Mt. Gox)
Craig List Rental Scams2
• Rental ad crawler
• Scam identifier based on keywords
• 9% clones of genuine ads
• 30% credit reference scams
Graph Analytics forTransaction Fraud Analysis3
• “X is paying Y“, no channel-specifics
• ABN Amro
• Shortes path algorithms to identify communities
• Classifier: random forest
References
• Conference programme: http://fc16.ifca.ai/
• Referenced papers:– 1 http://fc16.ifca.ai/preproceedings/33_Krombholz.pdf
– 2 http://fc16.ifca.ai/preproceedings/01_Park.pdf
– 3 http://fc16.ifca.ai/preproceedings/02_Molloy.pdf
Tracking Mobile Web Users Through Motion Sensors: Attacks and Defenses
• 30 devices. 10 samples per device.
• 25 features (mean, skewness, etc.) used.
• played no audio, sinus wave and a popular song.
• More phones added to data set reduce quality• Participate at
http://web.engr.illinois.edu/~das17/SensorDataCollection.html
The Price of Free:Privacy Leakage in Personalized Mobile In-App Ads
1) Are advertisements personalized?
2) What can an app learn about a user by observing personalized advertisements?
The Price of Free
• … more than 73% of ad impressions of 92% of users are demographically personalized.
• … predict a user’s gender, age and parental status is significantly higher than that of predicting other types of demographics information.
Centrally Banked Cryptocurrencies
• No double-spending
• Non-repudiable sealing
• Timed personal audits
• Universal audits
• Exposed inactivity
Equihash: Asymmetric Proof-of-Work Based on the Generalized Birthday Problem
Reference implementation of a proof-of-work
• 700 MB of RAM
• runs in 30 seconds on a 1.8 GHz CPU,
• increases the computations by the factor of 1000 if memory is halved, and
• presents a proof of just 120 bytes long.
Driller: Augmenting Fuzzing through Symbolic Execution
x = int(input())
if x > 10:
if x < 100:
print "You win!"
else:
print "You lose!"
else:
print "You lose!"
Let's fuzz it!
1 ⇒ "You lose!"
593 ⇒ "You lose!"
183 ⇒ "You lose!"
4 ⇒ "You lose!"
498 ⇒ "You lose!"
48⇒ "You win!"
x = int(input())
if x > 10:
if x^2 == 152399025:
print "You win!"
else:
print "You lose!"
else:
print "You lose!"
Let's fuzz it!
1 ⇒ "You lose!"
593 ⇒ "You lose!"
183 ⇒ "You lose!"
4 ⇒ "You lose!"
498 ⇒ "You lose!"
42 ⇒ "You lose!"
3 ⇒ "You lose!"
Driller
x = input()
if x >= 10:
if x % 1337 == 0:
print "You win!"
else:
print "You lose!"
else:
print "You lose!"
???
x < 10 x >= 10
x >= 10x % 1337 != 0
x >= 10x % 1337 == 0
Driller
x = input()
if x >= 10:
if x % 1337 == 0:
print "You win!"
else:
print "You lose!"
else:
print "You lose!"
???
x < 10 x >= 10
x >= 10x % 1337 != 0
x >= 10x % 1337 == 0
1337
Driller
Driller: Combining the Two
“Y”
“X”
Test Cases
“Cheap” fuzzing coverage
Tracing via Symbolic Execution
!
Control
Flow Graph
Reachable?
Driller: Combining the Two
“Y”
“X”
Test Cases
“Cheap” fuzzing coverage
Tracing via Symbolic Execution
“MAGIC”
New test cases generated
Control
Flow Graph
Synthesized!
Driller: Combining the Two
“Y”
“X”
Test Cases
“Cheap” fuzzing coverage
Tracing via Symbolic Execution
“MAGIC”
New test cases generated“MAGICY”
Control
Flow Graph
Towards completer code coverage!
32c3
Medial aufbereitet:
• Karsten Nohl: EC Kartenterminals
• NSA Untersuchungsausschuss
• VW “Dieselgate”
32c3
Wie immer:
• DECT/GSM, 40gbit, 4 parallele Tracks
• Neu: 20gbit DDoS von extern
Themen und Trends:
• Zensur (China, Iran, ...)
• akademisch++ (djb, Univ. of Michigan, ...)
Datahavens: From HavenCo to Today
Ryan Lackey (Cloudflare)
Sealand:
• WWII Waffenplattform in der Nordsee
• zuerst Wifi (PCMCIA), dann Mikrowelle
• 2000-2002 in Betrieb
Datahavens: From HavenCo to Today
Probleme:
• Bandbreite teuer (256kbit pro Server)
• Dotcom Blase geplatzt
• schlechte Organisation, kein Zahlungsdienstleister, einsam …
• keine Killer-Anwendung
• kein Bitcoin, keine Virtualisierung, ...
Rowhammer.js
Inhalt:
• März 2015: Project Zero baut 2 Exploits
• Erstmals in JavaScript gezeigt1
• häufiges Lesen von Speicher kann benachbarte Bereiche verändern
• Forscher von der TU Graz1 https://github.com/IAIK/rowhammerjs
Tor onion services
.onion Top Level Domain:
• Erreichbar über z.B. Tor Browser Bundle
• nicht nur Silk Road(s): Facebook, Wikileaks, SecureDrop, ProPublica
• Authentifiziert, verschlüsselt, anonym, NAT punching
Neither Snow Nor Rain Nor MITM
SPF, DKIM & DMARC vermessen1:
• STARTTLS in SMTP: MITM Angriff kann es unterdrücken
• Logs von > 1 Jahr Gmail
• Tunesien: 96% TLS stripping
• > 20%: Irak, Papua Neu Guinea, Nepal, Kenia, Uganda, ...
1 http://conferences2.sigcomm.org/imc/2015/papers/p27.pdf
Weitere Präsentationen
Weitere Präsentationen:
• Logjam1 (schwache DH-Gruppen, EXPORT Cipher)
• Lets Encrypt2
• DDoS Mitigation Fails
1 https://weakdh.org 2 https://letsencrypt.org