Coming soon — posts below are illustrative of planned writing topics.
As AI systems become increasingly autonomous, the alignment problem takes new dimensions. This post explores failure modes unique to agentic architectures and governance mechanisms needed to address them.
Can blockchain's immutability and decentralized verification solve the auditability crisis in autonomous AI? Reflections from our IEEE ICCA 2025 work on blockchain-monitored agentic pipelines.
Prompt injection is one of the most pervasive and underappreciated threats to LLM-based agentic systems. A deep dive into attack vectors, mitigation strategies, and open research questions.
Research articles on agentic AI safety are in progress. Follow on Google Scholar for publication updates.
Follow on Google Scholar