0%
Chat
← Back to all posts

A Communication Management Plan for the Age of AI

•5 min read•0 views•

The traditional communication management plan answers a familiar set of questions: who needs what information, in which format, how often, through which channel, and who is responsible. Those questions still matter. But they no longer cover the whole picture.

Today, a large share of project communication is touched by AI. Status updates are drafted by assistants, meetings are transcribed and summarized automatically, documents are translated in seconds, and dashboards generate their own narratives. That brings speed — and new risks: hallucinated facts in customer emails, unclear accountability, and records whose integrity is hard to prove.

This article outlines how to upgrade a PMBOK- and BABOK-aligned communication management plan for the age of AI.

What Stays the Same

The foundations of good communication planning remain:

  • Elicit stakeholder needs first. Following BABOK guidance, document what each stakeholder group needs before choosing channels.
  • Match format and frequency to the audience. Executives need concise, decision-oriented briefings; project teams need detail; regulators need precision and evidence.
  • Use the right mix of channels: interactive (meetings, workshops), push (emails, reports), and pull (dashboards, wikis).
  • Assign clear ownership for every communication.
Stakeholder Group Needs Format Frequency
Executive sponsors Progress, risks, decisions needed Formal, concise, data-backed Bi-weekly
Project team Status, blockers, working detail Informal, detailed Daily/weekly
Customers and clients Deliverables, impact, timelines Formal, plain language At milestones
Regulators and auditors Compliance status, controlled disclosures Formal, precise, evidenced As required
Public and media Announcements Formal, brand-reviewed As needed

Upgrade 1: Tiered Governance for AI-Assisted Communication

The core upgrade assigns every type of communication to a tier based on stakeholder impact and risk — not convenience.

Tier 1 — AI Permitted, Light-Touch Review

Low-risk, internal, reversible communications: first drafts of routine status updates and internal memos, meeting notes, summaries of long threads for internal use, translation of working documents, and brainstorming structure or talking points. A quick human read-through before sending is enough.

Tier 2 — Human Review Required

AI may assist, but a named reviewer must read, fact-check, and approve before release. This covers anything that reaches customers, executives, decision-makers, or the public — and any communication containing figures, dates, commitments, or claims.

Tier 3 — Restricted

Human subject-matter experts author or substantially rewrite these, and personally sign off: legal notices, contracts, regulatory filings, financial disclosures, payroll and benefits communications, HR decisions about employment or pay, and crisis and incident communications.

Upgrade 2: A Human-in-the-Loop Review Gate

Every Tier 2 and Tier 3 communication follows a fixed chain:

Draft → Fact-check against the source of record → Named reviewer → Accountable approver → Release

One rule underpins it all: no AI system is ever the approver of record. Accountability for every released communication rests with a named person.

Upgrade 3: An AI Disclosure Standard

  • Internal, Tier 1: no disclosure required, but AI assistance should be discoverable on request.
  • External, Tier 2 and 3: disclose material AI involvement where the audience would reasonably expect it — for example, "drafted with AI assistance, reviewed and approved by [name]" — in line with applicable regulation and brand policy.

Upgrade 4: Tamper-Evident Audit Trails

For communications with regulatory, financial, or evidentiary weight, integrity matters. A privacy-preserving model provides proof without exposing content:

  1. Store the final, approved communication only in an access-controlled system of record.
  2. At approval, generate a cryptographic hash of the final artifact, independently of the drafting tool.
  3. Retain the hash as immutable evidence that the record existed in that exact form at that time.
  4. Disclose the full record only under legitimate authority — a regulator, auditor, or court.

That gives tamper-evidence and chain of custody without replicating confidential or personal data. It's reserved for high-stakes communications; applying it to everyday messages would only add overhead.

Upgrade 5: Invest in Human Communication Skills

Here's the paradox of AI-assisted communication: when anyone can generate fluent text in seconds, fluency stops being a differentiator. What matters more is judgment, clarity, empathy, and persuasion — knowing what to say, to whom, and why.

Warren Buffett has often said that the communication course he took as a young man was one of the most valuable investments he ever made, and that brilliant ideas are worthless if you can't get others to follow them. That's even more true now.

The upgraded plan treats communication skill as a managed capability:

  • Skill tracks in clear writing, presenting, active listening, facilitation, and difficult conversations.
  • Development mechanisms such as workshops, peer review, coaching, and practice in real project settings.
  • Measurement through stakeholder feedback, review outcomes, and observed improvement over time.

Risks to Manage

Risk Mitigation
Hallucinated facts in external communications Tier 2 fact-checking against the source of record
Confidential data entered into AI tools An approved AI tool register and clear data rules
Unclear accountability Named reviewers and approvers for every Tier 2/3 item
Disputed records Hash-based integrity proofs for high-stakes communications
Skills atrophy A deliberate communication skill-building program

Conclusion

AI can make project communication faster, more consistent, and more accessible. But speed without governance creates risk, and fluency without judgment creates noise. An upgraded communication management plan keeps the proven PMBOK and BABOK foundations, adds clear rules for where AI may and may not be used, keeps humans accountable, protects the integrity of important records — and invests in the human skills that matter more than ever.