A General Manager — fifteen years of experience — sent her CEO a carefully worded email. Not a routine update. A candid observation about a strategic direction she believed was wrong, written with the kind of courage that only comes from years in the room. The reply came back in four minutes. Polished. Structured. And completely hollow.
She told me later: “I knew immediately. It wasn’t him. And I decided, right then, that I would never send him anything important again.”
That four-minute reply cost that CEO something no automation tool can restore — the candour of a senior leader who had something real to say.
When people write to you, they are not just sending information. They are sending courage, concern, and context. That deserves a human back.
We Have Confused the Tool With the Outcome
There is a quiet epidemic in India’s C-suites right now. CEOs, CMOs, and Business Heads are automating everything — calendars, email responses, WhatsApp replies, LinkedIn comments. Not always because it makes them more effective, but because it signals that they have “embraced AI.”
I use AI every day. I have built my practice around it. It compresses research, stress-tests strategy, and handles the groundwork that used to consume half a working week. That is exactly what it should do — clear the noise so the human can do what only humans can. But somewhere along the way, we started letting AI do the human part too.
Your team already has someone to crunch the data. They don’t need AI to read your emails. They need YOU.
What a GM’s Email to the CEO Actually Is
When a senior leader writes to you directly — especially about something difficult — that message carries weight far beyond its words. It carries professional risk. Emotional investment. It might be a warning you cannot afford to miss, a loyalty signal, or a distress signal. Often both at once.
No AI — not Claude, not Gemini, not GPT — can sense that. An AI reads the text. It cannot read the three drafts deleted before this one was sent. It cannot detect the hesitation behind a perfectly professional sentence. And when your AI replies to their humanity with a four-minute, well-structured, soulless response — you have not saved time. You have broken something.
The future we should fear is not AI taking our jobs. It is leaders voluntarily giving AI their most human responsibilities.
Where the Line Is
Automate ruthlessly below the line of human judgment — data, dashboards, summaries, first drafts, scheduling. AI will do all of it faster and cheaper. Let it.
But protect everything above that line. Strategic decisions. Difficult conversations. Any response to someone who took a professional risk to write to you. The test is simple: would the person feel respected knowing an AI wrote this? If the answer is anything but yes — write it yourself.
The damage from getting this wrong is invisible until it is catastrophic. Your team will not tell you they have stopped bringing you real concerns because they know Claude will answer. They will simply stop. The candour dries up. The early warnings go quiet. You clear the noise — and silence the signal.
AI is a productivity tool. The moment it becomes a relationship proxy, you have a leadership problem.
The leaders who win the next decade will not be the ones who automated the most. They will be the ones who automated the right things — and stayed fiercely, deliberately human where it mattered.
The world does not need more Gemini-to-Claude conversations. It needs more human ones.
