For Project Managers

20 Practical Ideas for Project Managers to Stay Cognitively Sovereign

AI project plans can look rigorous while hiding untested assumptions that senior PMs would spot immediately. When you outsource risk identification to templates, you lose the situational awareness that catches trouble early.

These are suggestions. Take what fits, leave the rest.

Download printable PDF

Reclaim Planning Judgement

Write your plan assumptions before asking AIbeginner
Document what you believe about timeline, scope, and team before Copilot generates structure.
Test AI plans against past project failuresintermediate
Compare generated schedule to three previous projects that went wrong. Note what AI missed.
Identify which milestones rely on untested dependenciesintermediate
Flag tasks where AI assumes handoffs, approvals, or third parties will cooperate on time.
Ask your team to challenge the AI draftbeginner
Run Monday.com plan past engineers and leads before stakeholders see it. Listen for silent nods.
Compare AI timeline to your gut estimate firstbeginner
Write your realistic schedule alone. Then compare to ChatGPT output. Notice where intuition diverges.
Spot resource levelling gaps in AI schedulesintermediate
Check if generated plan assigns one person to two critical tasks in the same week.
Track which assumptions your team actually questionsbeginner
In planning meetings, note where experienced people push back on AI suggestions. Record those concerns.
Separate AI-generated tasks from your project-specific onesbeginner
Create two lists: template tasks versus tasks only you know matter for this client.
Require manual approval before dependencies lock inintermediate
Do not let Jira AI auto-link tasks. Approve each dependency based on real team capability.
Document the reasoning behind your plan changesbeginner
When you override AI suggestions, write why in project notes. Build a record of your judgement.

Protect Risk and Status Signals

Build risk register from your team's lived experienceintermediate
Before using Notion AI template, interview three people who worked similar projects. Capture their fears.
Flag risks that AI templates always missintermediate
Add rows for your organisation's specific blockers: approval cycles, vendor flakiness, skill gaps, politics.
Test risk ratings against your escalation historyintermediate
Review past incidents that became crises. Check if AI-rated risks would have caught them early.
Record early warning signs before they become risksbeginner
When team morale drops or a stakeholder goes quiet, document the signal now. Do not wait for status reports.
Compare AI-polished status to your unfiltered notesintermediate
Write status summary in plain language first. Then check what Copilot changes. Ask why.
Require one sentence of evidence for each green flagbeginner
Do not let Notion AI mark tasks complete without proof. Insist on actual deliverable or test result.
Schedule unstructured check-ins with key team membersbeginner
Talk to engineers and designers outside of status reporting cadence. Feel the real mood. Listen for hesitation.
Keep a separate risk log that AI does not touchbeginner
Maintain a private notebook of concerns you do not yet want to escalate. Review it monthly yourself.
Audit stakeholder communication before sendingbeginner
Read your Monday.com report aloud before publishing. Check if it matches what you told the sponsor verbally.
Identify which status signals usually predict failureundefined
For your organisation type, list the red flags that historically meant a project would slip or derail.

Five things worth remembering

Related reads

The Book — Out Now

Cognitive Sovereignty: How To Think For Yourself When AI Thinks For You

Read the first chapter free.

No spam. Unsubscribe anytime.