Ground Truth·18 min
Preeti Ravindra
Shows how A/B experiments measure whether LLM tools actually help SOC analysts write incident findings faster and better.
aisecurity-operationsmachine-learning+2
Ground Truth BSidesLV 2024
Las Vegas, USA · August 2024 · 14 talks
Ground Truth talks from BSides Las Vegas 2024.
1–14 of 14 talks
Ground Truth·56 min
Fred Heiding · Simon Lermen
Shows how large language models can automate spear phishing campaigns and how personalized spam filters can counter them.
phishingaisocial+2
Ground Truth·42 min
Emanuel Valente
Shows how black-box adversarial attacks evade machine-learning models that detect DNS-over-HTTPS tunnels.
machine-learningaidetection+2
Ground Truth·37 min
Matthew Canham
Presents a cognitive attack taxonomy that links social engineering tactics against humans and AI systems.
socialaiattack+2
Ground Truth·21 min
Emily Austin
Shows how to count industrial control devices on the internet while filtering out honeypots, lab setups and false positives.
icsinfrastructureinternet+2
Ground Truth·34 min
Arun Vishwanath
Examines how parents decide when to give children smartphones, and where their cautious instincts may not match the evidence.
mobilerisklearning+2
Ground Truth·35 min
Suha Sabi Hussain
Shows how input-handling bugs in ML model file parsers can be chained into backdoors that carry through into deployed pipelines.
aimachine-learningvulnerabilities+2
Ground Truth·41 min
Brandon Pinzon
Shows how public SEC cyber filings can be paired with financial data to spot early warning signals of breach impact.
riskintelligencethreat+1
Ground Truth·45 min
Wendy Hou-Neely
Presents cyber insurance claims data on incident costs, ransomware payment trends, and privacy and business interruption exposure.
riskransomwareprivacy+3
Ground Truth·52 min
Bobby Filar
Shows how internet measurement studies can expose hack-for-hire phishing, scanner accuracy gaps, and patching friction.
phishingintelligencenetwork+2
Ground Truth·22 min
Hirofumi Kawauchi
Shows how a large Japanese telecom used SSVC decision trees to narrow 50,000 vulnerabilities to a realistic urgent set.
vulnerabilitiesriskdefensive-security+1
Ground Truth·41 min
Douglas McKee
Presents a method for reversing a proprietary protocol from packet captures, then applies it to attack and detection.
reverse-engineeringnetworkiot+2
Ground Truth·22 min
Ariana Mirian
Shows how quickly internet-wide scan results go stale, with uptime varying by port, protocol and network.
networkinfrastructureintelligence+1
Ground Truth·54 min
Ezz Tahoun
Shows how open-source language and graph models can group SOC alerts into ATT&CK-mapped attack campaigns.
security-operationsmachine-learningincident-response+2