31Agentic AI Security11 min read
Before a Marketing Agent Uses MCP, Draw the Blast Radius
A practical MCP threat model for agents that can read CRM data, browse untrusted pages, create campaigns, or call production tools.
Automation you can audit
AI Agents for Marketing · Page 2
AI agents for marketing are only worth running if someone can see what they did. These posts cover agentic workflows, approvals, logging, recovery and the work that should never vanish into a black box, plus when not to use an agent at all.
Pick by problem
Read the one closest to your decision.
Each article answers a different question. Start with the one that matches what is on your desk this week.
32Agentic AI Evaluation11 min read
Your Marketing Agent Passed the Demo. Now Give It an Eval.
A smooth demo says little about stale sources, conflicting CRM records, tool failures, or missing approvers. Turn those real conditions into the tests an agent must pass before it earns more authority.
33AI Workflow Governance10 min read
The Approval Gate Is a Product Decision, Not a Slack Message
Decide which AI-assisted actions may run, which must pause, what an approver needs to see, and how the workflow recovers when nobody answers the request.
34AI Data Governance11 min read
Do Not Send the Whole CRM to an AI Tool Because the Prompt Was Easier
A bounded AI task rarely needs an entire customer history. Map the minimum fields, every copy, each retention period, and the deletion path before connecting the workflow.
35Prompt Injection Defense11 min read
The Web Page Is Data. Your Agent May Mistake It for an Order.
How indirect prompt injection reaches marketing agents through pages, email, documents, CRM notes, and tool output—and how to contain it without pretending one system prompt is a firewall.
36Agent Permissions10 min read
Give the Agent the Smallest Key That Can Finish the Job
An instruction to “only update notes” is not access control. Separate read from write, scope records and fields, use short-lived credentials, approve material actions, and revoke stale grants.
37AI Auditability10 min read
If the Agent Changed the Campaign, the Log Should Explain More Than “Success”
What to record across prompts, sources, tool calls, approvals, state changes, cost, latency, and recovery so an AI-assisted marketing workflow can be debugged without creating a new data leak.
38AI Vendor Evaluation11 min read
Evaluate the AI Vendor With Your Workflow, Not Their Best Demo
The vendor’s best demo avoids your worst data and failure paths. Test representative work, data controls, reliability, operating cost, change management, and the effort required to leave.
39AI Incident Response11 min read
When the Marketing Agent Gets It Wrong, Stop the Side Effects First
An incident-response runbook for containing outbound messages, campaign changes, CRM writes, data exposure, and queued automations before investigating and safely restoring service.
40AI Operations10 min read
How to Design Governed AI Marketing Workflows
AI can shorten research, content, lead operations, and reporting work. This operating model shows where controls, review, traceability, and recovery still have to sit.
Start with a question
Ask these before you read.
- Which repeated work is safe to hand to an agent?
- Where must a person review, override or roll back?
- How will the team know the workflow helped?
The work behind it