Two Clocks Are Running. Only One of Them Gets Attention.

Elif Güvençer:

A committee convenes on AI adoption. The Chief Communication Officer finds out after the first meeting. A paper on AI risk goes from the Chief AI Officer to the board with no communications input, because nobody thought to ask. The CEO has started writing his own quotes and sending them over to comms for a look. The function that governed executive voice is now copy-editing it.

None of these is a crisis or a massive problem on their own but together, they mark a change in agency and mandate for communications. You can call it the beginning of the road to demotion.

It happens because the function is running two kinds of clocks at once, and only one of them gets attention. The immediate clock is outward-facing: monitoring, content, reputation management in an AI-mediated world. The structural clock is inward-facing and looks at what the function needs to become and how its value gets measured. It rarely gets any attention.

The immediate clock’s gravitational pull: AI engine reputation

Persuasion no longer starts from a blank page. It starts from a machine-made baseline.

Communications was built for a relational world. You can brief a stakeholder, correct a journalist, engage a regulator. You cannot do that with an AI engine. Not in a way that will change what it tells the next person who asks. So the first thing is to stop thinking about reputation management as a relational practice and start thinking about it as an architectural one. Something you architect one signal at a time.

Most of the effort right now goes to visibility: whether the organisation shows up when someone asks an AI engine. That is the wrong lens. The real question is eligibility: whether the organisation is credible enough to be selected at all. And that happens with what I call signal architecture: governing your reputation at the input level, creating and managing the signals on the surfaces the system ingests, rather than at the output level.

Signals need to be credible. This is something the old playbook of impressive brand claims misses. A confident line like “best car in the world” can move a human. A machine wants the evidence underneath it. It says don’t tell me, show me. Communications’ muscle of fact checking, claim finding and general commitment to truth makes it the ideal function to become this ‘truth layer’ that eligibility requires.

The total picture AI paints about an organisation has no default owner. Investor relations governs the financial narrative. Marketing governs brand. Legal reviews the disclosures. Communications governs media and executive voice. Each does its job, and optimises its individual surface. What nobody does is look at the total. That was survivable when nothing else could read the sum either. AI changes that because there is now an intermediary capable of reading the whole and handing the result to anyone who asks. This is the new governance gap communications needs to fill.

The risk horizon has to widen with it. What the organisation says does not fade into the background anymore after the news cycle is over. Any narrative debt, the gap between what you said before and what you say and do now, sits in the stretching memory of AI, available to anyone who asks. The one time you’d want AI’s memory problem to work in your favour, it doesn’t.

Reputation gets lost one token at a time. 

When something goes wrong, there is usually a trace. A decision was made, by someone, in a context you can reconstruct. AI breaks the chain. When the consequence originates inside the system, whether that is a fabricated answer or an agent operating outside its brief, reconstructing what happened and why is slow, and often impossible at the speed the situation demands. When everything accelerates, Communications’ role of bringing the trust consequence of a decision that looks fine from every other direction is needed more, not less.

This is also the argument for communications in AI governance. It rests on this capability of second order risk thinking and one more thing: the function’s independence. The Chief AI Officer is measured on adoption velocity. Marketing looks at brand. IT looks at deployment. Legal looks at exposure. Each reads AI through something else. Communications reads it to find out whether the total picture holds. A function with a stake in adoption cannot mark its own homework. 

There is a price of admission here. You cannot govern what you do not understand so leaders need AI literacy. Not necessarily enough to build the system but enough to challenge design and governance decisions while it is being built, not after it fails.

Whether you run it or not, the structural clock is ticking

AI is not a productivity tool. It is a structural stress test. It automates outputs the function has been measured on, leaving communications leaders with a hard question: where is my value proposition? Is it upstream or is it defined as the total sum of the volume and quality of what gets produced? Most will find the balance is not where they think.

After this honest audit, the next step is to define the strategic intent of the function. What do you exist to do in a world where AI is not just automating tasks but fundamentally changing how the system operates: the way we get information, make decisions and work? This intent needs to guide any redesign that follows and the skillset needed to carry the new structure.

When adopting AI, the focus should be on reclaiming bandwidth and redeploying it upstream. That means using AI strategically on what communications can do now and how its value gets measured. Not to do the same thing faster.

Mind the real tension that pushes the structural clock down the priority list

The reason for deferral runs deeper than bandwidth or time. Both clocks ask the leader to be professionally vulnerable, in two different directions at the same time.

The immediate clock asks them to show up as a risk and reputation thinker for machines, not just for people. Most have not publicly inhabited that profile yet, but it is a socially acceptable one to grow into. After all AI is new for everyone.

The structural clock asks them to take the function’s value proposition apart, visibly, under scrutiny, and rebuild it around what is needed now. And that is a vulnerability most will not choose.

That is how you get the picture at the top of this piece. Gradually, through decisions that look insignificant and justified at the time. If you’d rather avoid that destination, here is where to start:

• Write this sentence: this function exists to ___ in an AI-mediated world. Take it into your next three leadership conversations and see whether it survives contact.
• List the three tasks you would automate tomorrow. For each one, name where the reclaimed hours go. Efficiency via AI adoption is the byproduct, not the goal.
• Of the main skills needed: AI literacy, reputational risk applied to machines and coherence judgement for the picture AI engines paint of your organisation, define which ones you preserve, which you build and which you recruit.
• Take the three questions a buyer, an investor or a journalist would ask and ask them to the AI engine that your stakeholders use the most. The gap between what the engine says vs what you want it to say is your first signal architecture map.
• Find out where AI decisions are being made in your organisation and who is making them. If communications is not there, earn the entry on the second reading: whether the reputational picture is holding together, trust consequence of where it does not, read independently of everyone driving adoption.

Communications is the discipline built on anticipating consequence and thinking ahead of the problem. It rarely turns that instinct on itself. The most critical use of this capability is thinking about what is coming for the function itself.

+++

Elif Güvençer is a communications strategist, VP at RepTrak, and the author of The Two Clocks: A Framework for Communications in the Age of AI (reputationsignal.co).