
Compose an AI research boundaries memo
You'll have a 1-page internal memo defining what your team will and won't use AI research agents for, including compliance constraints, required human review steps, and quality thresholds.
What you'll make
AI Research Boundaries Memo
- Approved Use Cases: what we'll trust AI research agents for
- Required Human Review: where human expertise must verify AI output
- Quality Thresholds: how we measure when AI research meets our standards
- Off-Limits Scenarios: what we won't use AI research for and why
Write as an internal memo your team lead could send tomorrow — specific enough to guide daily decisions
The 4 parts
Each part ends with a check you can see.
Map Your Research Landscape
Catalog your team's actual research tasks and identify where quality matters most versus where speed wins.
Done when
You have a ranked list of 8-12 research tasks your team does, scored by quality-sensitivity and time pressure, with notes on current verification methods.
Test Against Real Standards
Run an AI research agent on one of your actual research questions and evaluate where it meets versus misses your quality bar.
Done when
You have tested one multi-step research question, documented 3+ specific verification points where you had to override or supplement AI output, and identified your team's actual quality threshold.
Design Human Checkpoints
Define the specific moments where humans must step in to verify, redirect, or override AI research output.
Done when
You have a 3-5 step handoff protocol specifying when humans review AI research, what they check for, and how they document overrides.
Draft the Boundaries Memo
Synthesize your findings into a one-page memo that gives your team clear guidance on AI research use.
Done when
You have a complete 1-page memo with specific use cases, quality gates, and scenarios that any team member could reference for daily research decisions.
Capstone: you put together your AI Research Boundaries Memo.
Where this helps
An elective: a single lesson that sits outside rtwo's job certificates and doesn't count toward one.
About this lesson
What you'll be able to do
- Draft a 1-page research boundaries memo that your team can actually implement tomorrow
- Identify 3-5 specific handoff points where human oversight prevents research blind spots
- Test one AI research workflow against your team's actual quality standards and document what passes vs fails
Who it's for
- Team leads who need to set guardrails for AI research tools
- Anyone responsible for research quality standards in their organization
- People who've tried AI research and want structured boundaries rather than ad-hoc decisions
What it won't cover
- Legal compliance review processes — this assumes you know your regulatory constraints
- Budget allocation for AI research tools
- Training non-technical staff to use research agents
Skills you'll practise
- Research quality threshold calibration
- Human-AI handoff design
- Source verification workflows
- Risk boundary articulation