Lanthos

Report / 7 chapters

AI Governance & Agent Risk

A practitioner's report on delegation boundaries, prompt injection as a capability problem, connector risk, browser agents holding live sessions, and how EU AI Act timing lands on operations.

What's inside
  • Treating shadow AI as a requirements document instead of a violation
  • Why every connector you approve is a delegation decision
  • What changes the moment an agent can spend money
  • Reading the EU AI Act timeline as an operating constraint, not a legal one
Chapters
  1. 1

    Shadow AI is a requirements document

    11 min read

  2. 2

    The deadline moved — your exposure didn't

    11 min read

  3. 3

    AI browser agents inherit everything

    10 min read

  4. 4

    Prompt injection is not a bug

    5 min read

  5. 5

    Integration was the checkpoint

    5 min read

  6. 6

    An agent with a wallet

    11 min read

  7. 7

    The double-edged arms race

    13 min read

Written for: Risk, compliance, security, and AI governance leaders.