Where AI Actually Helps (and Where It Doesn’t)
Communicators already know this truth:
Not everything should be automated.
And not everything can be.
AI is powerful, but it’s also blunt. It accelerates the parts of our work that are mechanical — and complicates the parts that rely on judgment, timing, and trust.
Over the past few months, I’ve been doing foundational AI training. What struck me is how little of it speaks to the realities of communications work — the judgment, the nuance, the trust. That gap is exactly why I’m building this toolkit.
Because the real skill isn’t “using AI more.”
It’s knowing where AI belongs in the workflow… and where it absolutely doesn’t.
Where AI Actually Helps
1. Sensemaking before writing
AI is excellent at compressing raw information:
meeting transcripts
long documents
feedback threads
survey comments
It gives you a starting point so you can focus on meaning, not extraction.
2. Drafting early passes (not final messages)
AI can:
outline
structure
suggest framing
flag missing context
It gets you out of the blank‑page stage so you can stay in editor mode — where your judgment lives.
3. Adapting approved content across channels
Once the message is set, AI can help you:
shorten for Slack
expand for intranet
simplify for managers
adjust cadence and tone
It keeps consistency without forcing you to rewrite the same thing five times.
4. Acting as a second set of eyes
AI is surprisingly good at:
spotting jargon
flagging tone drift
checking clarity
identifying likely audience questions
It’s friction, not replacement.
Where AI Doesn’t Help (and often makes things worse)
1. Anything requiring human accountability
Apologies, layoffs, harm, fear, grief — these moments need a human voice.
AI can’t carry relational weight.
2. Situations where ambiguity is intentional
Sometimes “we don’t know yet” is the message.
AI tries to fill gaps that shouldn’t be filled.
3. Messages rooted in history, culture, or unwritten context
AI doesn’t understand:
inside jokes
past promises
political landmines
who gets the benefit of the doubt
It can sound “technically fine” but socially wrong.
4. External‑facing interpretation
AI can prepare communication.
It should not be the communication.
If a mistake would damage trust or reputation, AI belongs far from the final output.
The Real Skill: Discernment
Before using AI, ask:
Is this task mechanical or meaningful?
Mechanical → AI can help
Meaningful → Humans lead
Is the risk confusion or credibility?
Confusion → AI can be closer
Credibility → AI stays far away
Does the audience expect a human?
If yes → AI supports, not speaks
My Quick Win
AI can speed up the work, but only discernment protects it.
The win isn’t using AI everywhere — it’s knowing exactly where it belongs.
