Journal · ABDC A
PhD · Information Systems · UNT
Ali Safari.
I study how people and organizations can trust AI when the stakes are real: security, governance, and everyday work.
- 10
- papers
- 10+
- years in tech
- 345
- field notes
Agentic AI
Trust
Cybersecurity
Ideas that have to work in organizations. Not just on slides.
About
As a PhD candidate at UNT, I work with Dr. Dan J. Kim on trust and security, Dr. Justin Ku on open-set recognition, Dr. Anna Sidorova on agentic AI delegation, and Dr. Hoon Choi on social media analytics.
Before academia I built technology businesses for a decade, including Rasa. During the PhD I also shipped RaSEC, a free security platform for students and researchers.
Teaching
BCIS 3610.008
Basic Information Systems · Mon 6:35–9:25 PM · BLB 070
Program requirements →
Research
Six threads I keep pulling.
01
Agentic AI & Delegation
How humans and AI systems hand off authority, and how control stays sane when the agent moves fast.
Explore → 02Trust in Intelligent Systems
When trust is calibrated, overdone, or broken. Especially in cybersecurity decisions.
Explore → 03Open-Set Threat Detection
Teaching models to notice attacks they have never seen, with Extreme Value Machines.
Paper → 04AI Governance
Who is accountable when an AI system fails, and what organizations actually put in place.
Explore → 05Privacy & Disclosure
Why people share data in one context and freeze in another. Risk, benefit, boundary.
Explore → 06IS Theory & Adoption
How organizations absorb new systems. Classic IS theory meeting messy practice.
Explore →AMCIS · CORE A
Open-Set Fake News Detection Framework: Addressing Algorithm Aversion through AI Geometric Bounding
AMCIS ↗AMCIS · CORE A
Managerial Discretion as Operational Insurance: Quasi-Experimental Evidence from Professional Football
AMCIS ↗ICIS · CORE A*
Enhancing IoT Security and Information Systems Resilience: An Extreme Value Machine Approach
ICIS ↗ICIS · CORE A*
Decomposing Attention in Creator Platforms: Content Quality, Social Capital, and Algorithmic Exposure
ICIS ↗ASC
Calibrating Human Trust in AI Cybersecurity Alerts: Real-Time Confidence Feedback
No items in this filter.
Field Notes
345 short reactions
Quick takes on IS research, AI, and theory. Not tidy summaries.
Comps
The IT Artifact Is Not Optional
Fifty-six percent. Sarker et al. coded the IS literature and found that is how often the IT artifact itself was missing.
Read → Agentic AIAgentic AI Changes Organizational Design
I counted eleven posts on agentic AI market sizing. Almost none asked what the organization has to become.
Read → TheoryThe Cognitive Offloading Problem
Five hundred and ten citations in its first year. And IS still barely asks what we stop practicing when AI remembers for us.
Read →How I see the field
Integration over adoption. Decision quality over prediction theater. Security and privacy in the center, not the appendix.