Reusing a Human Login Cannot Govern an Agent's Authority
Instead of reusing a human session, separate the principal from the agent and evaluate a delegation that binds resource, action, audience and lifetime at execution time.
Read articleRESEARCH TOPIC / SAFETY-RESILIENCE
Safety & Resilience
Attacks, incidents, recovery, and business continuity.
GROUNDING MAP
Questions and evidence in this topic
Explore 9 research notes through a shared decision domain. Each article identifies its conclusion, sources, findings, and limitations.
Related research notes
09 ARTICLES
Increasing the number of warnings does not necessarily increase safety.
Given habituation to repeated warnings, false alarms, and ignoring due to irrelevant notifications, design a warning budget by risk, display variations, and mandatory stops.
Read article
Even if you entrust your thinking to AI, can you return to it on your own?
We distinguish between load reduction via external memory and the reduction of internal encoding and retrieval, measuring both the performance when AI is usable and the independent execution when it is not.
Read article
The knowledge base of RAG becomes a searchable attack surface.
We mitigate the risk of malicious documents entering the top search results and guiding answers through write history, index monitoring, contradiction detection, post-generation verification, and isolation with re-indexing.
Read article
Code written by AI must be reviewed for speed and safety separately.
Do not equate the productivity of generated code with safety; instead, treat vulnerabilities, user safety perception, static and dynamic testing, and independent review as gates for adopting high-impact code.
Read article
Prompt-injection defense starts with permission boundaries, not instruction detection
Filtering suspicious text alone cannot protect AI that reads external data and executes tools; we organize multi-layered defense through attack evaluation, structured input, and official classification.
Read article
AI vendor selection must test the ability to resume operations after exit
We organize vendor evaluations that test data, model dependency, contracts, portability, and stoppage/alternatives before implementation, rather than selecting based solely on accuracy demos.
Read article
AI incident response does not end with system recovery
We outline an incident design that connects damage assessment, causality, evidence of versions/data/decisions, external reporting, and recurrence prevention, rather than ending with model suspension alone.
Read article
What Is Lost in Organizations That Rely Too Heavily on AI
Proper support helps; flawed support amplifies errors. Drawing on research into automation bias and critical thinking, we consider how to build organizations that can question, stop, and revert.
Read article