Latest posts tagged with #PaperPreview on Bluesky
The final #ACSAC2024 #PaperPreview today is Fan et al.'s work. While #crypto functions are essential but often mishandled, the authors evaluate crypto detection tools, highlighting gaps & paving the way for future work: www.openconf.org/acs...
#cybersecurity
First up as #ACSAC2024 #PaperPreview today is by Bao & Bindschaedler who analyze #reproducibility of SoTA #DPML methods. Findings are mixed. They discuss challenges and suggest best practices for future research: www.openconf.org/acs...
#cybersecurity #ML #deeplearning
Tao et al.'s "Exploring Inherent Backdoors in Deep Learning Models"
The third #ACSAC2024 #PaperPreview today is Tao et al.'s work shich shows #backdoor #attacks in #machinelearning can be realized without poisoning the data or model, referred to as inherent backdoors: www.openconf.org/acs...
#AI #ML #cybersecurity
Weber et al.'s "No Leakage Without State Change: Repurposing Configurable Cpu Exceptions To Prevent Microarchitectural Attacks"
The first #ACSAC2024 #PaperPreview for today is Weber et al.'s work, which shows that existing HW features can be repurposed to stop ongoing #SideChannel #attacks almost immediately by monitoring the victim: www.openconf.org/acs...
#cybersecurity
Fu & Wang's "Leaky Autofill: An Empirical Study On The Privacy Threat of Password Managers' Autofill Functionality"
The third #ACSAC2024 #PaperPreview for today is Fu & Wang's work in which they show that most #PasswordManagers inadvertently #autofill sensitive data into hidden field, revealing more serious privacy threat than previously known: www.openconf.org/acs...
#cybersecurity
The second #ACSAC2024 #PaperPreview today is Warnecke et al.'s work in which they implant an #ML #backdoor within a hardware accelerator, bypassing all software defenses with no run-time overhead: www.openconf.org/acs...
Zhang et al.'s "Stealing Watermarks of Large Language Models Via Mixed Integer Programming"
Today's first #ACSAC2024 #PaperPreview is Zhang et al.'s work in which they model the LLM watermark stealing attack as a constrained optimization problem to steal the details of the watermark and remove them: www.openconf.org/acs...
He final #ACSAC2024 #PaperPreview for today is Kleftogiorgos et al.'s Sidecar, a novel approach to create secure append-only channels using debugging logging, decoupling security checks to a monitor in a separate process: www.openconf.org/acs...
Kersten et al.'s "A Security Alert Investigation Tool Supporting Tier 1 Analysts in Contextualizing and Understanding Network Security Events"
Up next as second #ACSAC2024 #PaperPreview for today is Kersten et al.'s work on an alert investigation support system, a widget on top of a known SIEM which aids analysts into collecting the right information: www.openconf.org/acs...
@securescientist.bsky.social
The first #ACSAC2024 #PaperPreview today is Draschbacher & Maar's work uncovering serious flaws in Google's scheme for ensuring the integrity of apps published in the new AAB distribution format: www.openconf.org/acs...
Shen & Jin's "Reading It like an Open Book: Single-trace Blind Side-channel Attacks on Garbled Circuit Frameworks"
The third #ACSAC2024 #PaperPreview for today is Shen & Jin's work showing that side-channel attacks on garbled circuits reveal private inputs by exploiting secret-dependent AES sequences with a high success rate: www.openconf.org/acs...
Ferens et al.'s "Securing PUFs via a Predictive Adversarial ML System by Modeling of Attackers"
The second #ACSAC2024 #PaperPreview for today is Ferens et al.'s work, tackling the problem of secure and low-cost authentication in, e.g., IoT settings, by leveraging PUFs with adversarial ML: www.openconf.org/acs...
Today's first #ACSAC2024 #PaperPreview is by Ahmad et al., whose work "Eunomia" is the first real-time privacy compliance firewall for Alexa skills. Eunomia hijacks, examines, and validates skill communications: www.openconf.org/acs...
The second #ACSAC2024 #PaperPreview today is Mischinger et al.'s IoC Stalker whic detects underground forum posts containing malicious IoCs from the context ~355 days before they appear in OSINT: www.openconf.org/acs...
The first #ACSAC2024 #PaperPreview for today is Yilmaz et al.'s work which outlines the impact of #DDoS in modern #warfare, highlighting the importance of better protecting cyber infrastructure: www.openconf.org/acs...
#cyberwar #cybersecurity
Today's third #ACSAC2024 #PaperPreview is the work by Doan et al. who show “Digital” data poisoning is ineffective in the real-world based on a new DriveByFlyBy real-world test data set of #backdoor attacks in the wild: www.openconf.org/acs...
#cybersecurity
Xia & Chen's "Moire Injection Attack(MIA): Compromising Autonomous Vehicle Safety via Exploiting Camera's Color Filter Array (CFA) to Inject Hidden Traffic Sign"
The second #ACSAC2024 #PaperPreview for today is Xia & Chen's work showing a new threat to #AutonomousVehicles. MIA attacks inject invisible images into cameras, fooling #AI & risking safety: www.openconf.org/acs...
An online demo is available: sites.google.com/vie...
Kondracki et al.'s "Ready Or Not, Here I Come: Characterizing The Security of Prematurely-Public Web Applications"
Our first #ACSAC2024 #PaperPreview today is by Kondracki et al. who show that many #servers start out insecure and only become #secure in the hours and days after they become publicly accessible: www.openconf.org/acs...
#websecurity #TLS
@nicknikiforakis @briankon116 @sbucompsc
The second #ACSAC2024 #PaperPreview for today is the work by Armknecht et al. who devise a novel and efficient light client protocol that optimizes for the fact most users are rarely offline for more than a week: www.openconf.org/acs...
#Blockchain #Committee
The first #ACSAC2024 #PaperPreview today is Alla et al.'s work which proposes a flexible and seamless smart #authentication approach based on hardware impairments of devices: www.openconf.org/acs...
#ML #PLA #Exploiting #Hardware
Benita et al.'s "Vaktble: A Benevolent Man-In-The-Middle Bridge To Guard Against Malevolent Ble Connections"
The second #ACSAC2024 #PaperPreview today is Benita et al.'s VaktBLE, a friendly #MitM bridge to validate #BLE packets and block over-the-air attacks in real time: www.openconf.org/acs...
#cybersecurity #bluetooth
@GeovaniBenita @MatheusGarbelin @chatsudi
The first #ACSAC2024 #PaperPreview today is Ayoub et al.'s Bluescream, where they investigate how an attacker can manipulate #Bluetooth Low Energy to enable the long-distance EM #sidechannel #attack called Screaming Channels: www.openconf.org/acs...
#potatosecurity
The first #ACSAC2024 #PaperPreview today is Ayoub et al.'s Bluescream, where they investigate how an attacker can manipulate #Bluetooth Low Energy to enable the long-distance EM #sidechannel #attack called Screaming Channels: www.openconf.org/acs...
#cybersecurity
Ahsan et al.'s "Biosafe: Bioprinting Security Framework For Detecting Sabotage Attacks On Printability and Cell Viability"
The final #ACSAC2024 #PaperPreview today is Ahsan et al.'s work BioSaFe, which discusses #sabotage attacks on 3D #bioprinters using live cells and proposes a framework to detect such #attacks: www.openconf.org/acs...
#cybersecurity
Next up as #ACSAC2024 #PaperPreview today is Achamyeleh et al.'s "A Fly On The Wall - Exploiting Acoustic #SideChannels In Differential Pressure Sensors" showing how differential pressure sensors can capture speech, exposing a new #privacy #vulnerability: www.openconf.org/acs...
Lit et al.'s "Fedcap: Robust Federated Learning Via Customized Aggregation and Personalization"
The first #ACSAC2024 #PaperPreview today is Li et al.'s FedCAP, is a robust FL framework against both data heterogeneity and Byzantine attacks through customized aggregation and personalization:
www.openconf.org/acs...
#AI #FederatedLearning #Cybersecurity
Wang et al.'s "Physical ID-Transfer Attacks against Multi-Object Tracking via Adversarial Trajectory"
The second #ACSAC2024 #PaperPreview today is Wang et al.'s work about a new #vulnerability revealed in multi-object tracking used in #AutonomousVehicles and #surveillance where IDs can be fooled by physical maneuvers: www.openconf.org/acs...
#cybersecurity
Bhusal et al.'s "SECURE: Benchmarking Large Language Models for Cybersecurity"
The first #ACSAC2024 #PaperPreview today is Bhusal et al.'s work in which they propose 6 benchmark tasks to evaluate #LLMs in a #cybersecurity industry, provide insights & offer recommendations for improving LLM reliability: www.openconf.org/acs...
#AI
The third #ACSAC2024 #PaperPreview today is Chen et al.'s CubeVisor, a secure VM arch based on ARM CCA, which instantiates RMM and provides one-to-one service for each VM, protecting against other VMs: www.openconf.org/acs...
#potatosecurity #virtualisation
Chen et al.'s "CubeVisor: A Multi-realm Architecture Design for Running VM with ARM CCA"
The third #ACSAC2024 #PaperPreview today is Chen et al.'s CubeVisor, a secure VM arch based on ARM CCA, which instantiates RMM and provides one-to-one service for each VM, protecting against other VMs: www.openconf.org/acs...
#cybersecurity #virtualisation
