2026 • Forensic Science International: Digital Investigation
With or without logs: Memory forensic reconstruction of Model Context
Protocol (MCP) activity in agentic LLM systems
A. Satter, M. Salmon, L. Muhanna, T. T. Spinosa,
T. Gharaibeh, I. Baggili
Reconstructs Model Context Protocol activity from memory in agentic LLM
systems, showing how tool-use behaviour and system interactions remain
recoverable even when no logs were kept.
FSI: Digital Investigation, vol. 57, art. 302130
Read paper
2026 • DFRWS EU
Ex Machina: A forensic evaluation of AI companion applications and their
evidentiary value
K. J. Comeaux, T. T. Spinosa, A. Ghosn, I. Baggili
First cross-application forensic study of AI companion platforms totalling
over 25M downloads, with artifact-based user profiling and an assessment of
evidentiary value.
FSI: Digital Investigation, vol. 56, art. 302050
Read paper
2025 • ARES / WSDF
Money on My Mind: Forensic Investigation of Venmo Payment App
T. Spinosa, A. Onik, J. Rovira, I. Baggili
First-author study of vulnerabilities in Venmo through network traffic
inspection and logical acquisition, including MFA bypass pathways and API
session-token abuse for transaction recovery. Released as the PyMent tool.
Int. Conf. on Availability, Reliability and Security, pp. 60–77
Read paper
2024 • DFRWS USA
Hit and run: Forensic vehicle event reconstruction through driver-based cloud
data from Progressive's Snapshot application
A. R. Onik, T. T. Spinosa, A. M. Asad, I. Baggili
Extracts and validates Progressive Snapshot artifacts, demonstrating accurate
speed and location recording and revealing driving behaviour not exposed in
the mobile interface. Released as the PyShot tool.
FSI: Digital Investigation, vol. 49, art. 301762
Read paper