We often talk about AI as a tool that amplifies communication.
But we rarely ask: What kind of communication is it amplifying?

At AICommLab, we’ve spent the past few months exploring not just how AI communicates, but what that communication means—for design, for society, and for the future of agency.

In June, that conversation took on a more human dimension: mental health.

Not because AI has emotions (it doesn’t).
But because AI is increasingly shaping the emotional infrastructure of our interactions.

And that should concern anyone working at the intersection of technology and communication.

Systems Built for Speed, Not Substance

The digital ecosystem has taught us that success comes from clarity, speed, and virality.
This isn’t inherently bad—but it is incomplete.

When we teach AI to optimize for engagement, it learns to:

• Prioritize what’s already popular
• Polish away emotional nuance
• Avoid uncertainty or contradiction
• Simulate empathy without creating space for it

The result?
AI that communicates fluently, but says very little.
AI that mirrors our culture’s worst communicative habits.

The Mental Health Layer

Why does this matter for mental health?

Because we’ve built systems that reward performance over vulnerability.

And in those systems, certain voices—especially men’s—are taught to stay silent.
Not because they don’t feel.
But because the format doesn’t reward their truth.

This is the paradox we must design against:
AI can simulate conversation, but it can also normalize the absence of real dialogue.

If our digital tools echo only what’s optimized, then voices that are raw, contradictory, or slow to form… get silenced.

Toward a Different Design Ethic

This isn’t a call to make AI more emotional.
It’s a call to make the systems surrounding AI more accountable to emotion.

At AICommLab, we believe:

• Interfaces shape interaction
• Optimization isn’t neutral
• Format is never just aesthetic—it’s ethical

So our research asks: What would it take to design AI systems that make space for unformatted, unoptimized, and even unfinished expressions?

Because until we answer that, AI will keep learning from our silence.
And that’s not the legacy we want to leave behind.

Coming soon: Our new explainer video series on AI x Communication Systems.
Subscribe for updates, research insights, and student-led perspectives.