Upcoming research topic
Can Root-Privileged AI Agents Really Protect Your Files?
Examining how AI agents with elevated permissions interact with file and kernel security boundaries.
Cybersecurity research blog
Exploring the Security Boundary
Linux systems, AI agent security, storage security, eBPF, and hands-on research notes.
zedong@research:~$ cat research_interests.txt
Linux Kernel · eBPF · AI Agent Security · Storage Systems
● LINUX KERNELeBPF / LSMDATA INTEGRITY
// LINUX // SECURITY // SYSTEMS // OPEN RESEARCH
Research Channels
Different angles. A more secure computing world.
RESEARCH RADAR / SIGNAL ACTIVELIVE CFP FEED · ACADEMIC TRACKER
AUTOMATED INTELLIGENCE · SUBMISSION TRACKER
Launch full CFP operations center ↗
Security Conference Deadline Radar
A neon-lit radar for upcoming paper deadlines. Live countdowns, original/local time zone checks, and research-grade CFP references — with official dates clearly separated from community listings.
> NEXT SUBMISSION SIGNAL
UPCOMING MILESTONES / NEXT 03 SIGNALS
CT-RSA 2027 CCF C
Main deadline · paper Community · check CFPLoading time zone…
IEEE HOST 2027 NOT MAPPED
Main deadline · paper Community · check CFPLoading time zone…
IEEE S&P 2027 CCF A
Cycle 2 · abstract Official-checkedLoading time zone…
IEEE S&P 2027 CCF A
Cycle 2 · paper Official-checkedLoading time zone…
PETS / PoPETs 2027 CCF C
Cycle 1 · paper Community · check CFPLoading time zone…
Euro S&P 2027 CCF C
Main deadline · paper Community · check CFPLoading time zone…
AsiaCCS 2027 CCF C
Main deadline · paper Community · check CFPLoading time zone…
CHES / TCHES 2027 CCF B
Cycle 2 · paper Community · check CFPLoading time zone…
CCF ranking labels reference the CCF Network & Information Security catalogue ↗. They are distinct from CFP verification. Community listings are preliminary; verify all submission dates with the official CFP. Open the operations center for CCF A/B/C filters, research categories, a full Timeline, and calendar reminders.
FEATURED INVESTIGATION · SYSTEMS SECURITY
When AI Agents Go Rogue:
When AI Agents Go Rogue:
Who Protects Your Files?
An empirical investigation into the limits of host-side data protection when an AI agent has root privileges. Explore the path from agent actions through Linux security controls to persistent storage.
root@research-lab~/threat-model
$ investigate --layer lsm
$ trace --target block-io
$ evaluate --boundary storage
RESEARCH QUESTIONS / NO LIVE SYSTEM ACCESS
01 /Privilege & AgencyWhat changes when the agent runs as root?
02 /Kernel PolicyLSM enforcement, AppArmor, eBPF hooks.
03 /I/O ObservabilityFile operations and storage block traces.
04 /Trust BoundariesSeparating host control from data integrity.
Latest Posts
Research, experimentation, and technical deep dives.
The first blog articles are being prepared. Here are the research themes I plan to write about.
Upcoming research topic
Tracing Block I/O in Linux: Practical Notes
A technical walkthrough of local storage I/O, observability, and trace collection in Linux.
Upcoming research topic
eBPF LSM vs AppArmor: What Actually Happens?
Comparing kernel security controls and their behavior under privileged workloads.
No matching articles. Try another keyword.
GLOBAL RESEARCH VISITOR RADAR
● PRIVACY-FIRST TELEMETRY
Visitors Around the World ↗
An approximate geographic distribution of visitors to my research website. Individual IP addresses are never published.
Visits · 90 days—
Countries / regions—
Approximate locations—
Visitor analytics are not connected yet. No visitor locations have been recorded by this feature.
Opt-in only. No full IP addresses or GPS coordinates are saved or shown. Anonymous approximate locations are kept for up to 90 days.