Authenticity, Perspective, and Ethics in AI-Driven Media: What PR Can Use Today
At a recent industry event, Prasoon Joshi delivered a blunt truth: AI can scale creativity, but it can't replace the human instinct behind it. For PR and communications leaders, that's the line in the sand. Use the tool. Keep the soul.
What follows isn't theory. It's a practical filter for your campaigns, crisis playbooks, and AI policy-rooted in authenticity, perspective, and values.
Authenticity over amplification
Joshi traced his craft to a childhood moment of unspoken fear-a feeling that later became raw material for his writing. The point: real work starts where data ends. Authenticity isn't a tone. It's the decision to say what's emotionally true, not just what's popular.
He recalled Pandit Jasraj's approach: pick one person in the audience and perform for them. Try that in PR. Write to an actual human-your sister, a skeptical journalist, a friend who'd call you out. If it rings true for one, it scales to many.
- Practical move: Build a "one-person brief." Name the person. List their fears, hopes, and bias. Write to them. Then check if the copy still holds its center after approvals.
- Quality check: Ask, "What's the unspoken truth here?" If you can't answer, you don't have a story yet-you have copy.
For deeper, ongoing coverage of this topic, explore AI for PR & Communications.
Censorship pressures in the social media minute
As former CBFC chair, Joshi saw how judgment now lands in minutes. A film releases; the internet tries it in real time. One piece of content can trigger cross-border reactions before lunch.
His stance: replace confrontation with conversation. Most creators aren't trying to offend. Conflicts often come from vantage points. A river looks aggressive to the swimmer and calm to the bystander. Same river, different view.
- For PR teams: Move from "argue" to "align." Map the opposing view in one paragraph. If you can state it fairly, you can address it precisely.
- Response window: Draft tiered statements for the first 30 minutes, 3 hours, and 24 hours. Publish the smallest truth fast; update as facts firm up.
- Cross-border lens: Pre-brief local advisors in sensitive markets. A quiet WhatsApp with context beats a loud correction later.
AI can remix the past. Humans name the new.
Joshi doesn't see authorship as absolute-ideas pass through us. And AI isn't fully "artificial." It's trained on us. So yes, it will reorganize what we've already said and make it sound polished.
The line is this: genuine creativity often starts in the unexpressed. It shows up as intuition first, language later. If your work is just structural permutation, AI will catch it. If it's a leap, it stays human.
- Team policy: Use AI for drafts, research, and variants. Keep humans on the first spark and the final say.
- Novelty test: "What here would make a thoughtful critic pause?" If nothing, you're playing inside the mold.
- IP + consent: Log prompts, sources, and model versions. If it influences public opinion, document how it was made.
Formats aren't shrinking. They're splitting.
Shorts trend. Long form still wins attention when trust is earned. That's not new. People have always mixed snippets and depth based on need and timing.
Your job is to match intent to format, not chase whatever's trending right now.
- Two-track plan: Short hits for awareness and speed. Deep dives for authority and search equity. Publish both on a cadence.
- Signal > noise: Measure saves, replies, average watch time, brand search, and journalist pickups-not just views.
Break the mold or become the mold
Joshi's sharpest critique: AI often produces the "lowest common denominator of exceptional." It sounds profound because it matches what culture already agrees with. Safe. Familiar.
Human originality breaks the pattern. AI works inside it. Once AI outputs a pattern, it cements the norm for the next round. Your edge is to step outside that circle.
- Cliché audit: Kill lines you've seen thrice this month. If a sentence could sit in any competitor's post, it doesn't belong in yours.
- Specificity rule: Add one concrete detail that only your brand could say. If legal removes it, replace it with a different specific-not a generality.
The real debate: values, not features
Tech won't decide where the line is. People will. With AI agents, real-time targeting, and hyper-personal content, the question is simple: who sets the boundary, and why?
Integrity scales slower than manipulation, but it compounds. As Joshi noted, trust rooted in conviction beats clever tactics. You feel that difference in the work, and so does the public.
- Set your red lines: No synthetic endorsements. No deepfakes of real people without explicit consent. No dark patterns in prompts or targeting.
- Human in the loop: A named reviewer signs off on sensitive outputs. Not a team alias-a person.
- Explainability: If you can't explain how a message was produced or targeted, don't ship it.
- External anchors: Pressure-test policy against global standards like the UNESCO Recommendation on the Ethics of AI and the EU AI Act overview.
What PR can do this week
- Run a one-hour "unspoken truth" workshop for your flagship narrative. Capture the line that scares you a little. That's the seed.
- Ship a barbell content sprint: three short insights and one 1,200-word explainer. Track saves and replies, not just reach.
- Publish a public AI use note on your newsroom page: what you automate, what stays human, and how you protect audiences.
- Rehearse a 30-minute crisis drill using a recent cultural flashpoint. Aim for clarity, not clever.
If you want structured training to operationalize this, start with the AI Learning Path for Public Relations Specialists.
Closing thought
Anxiety drops when you zoom out. You're part of a much larger creative order. Use AI to scale the mechanics. Guard the instinct. Break the mold when it matters, and invest in conscience before you need it.
Your membership also unlocks:
