AI and Advertising Photography: What Changes, What Doesn't

Six months ago a beauty client sent me a folder of AI-generated images and asked, in the politest possible way, whether they still needed to book a shoot day at all. The images were seductive on a phone screen — flawless skin, impossible light, a model who didn't exist. Then we put them on a 6-sheet mockup and zoomed to 100%, and the whole thing fell apart: hair that melted into the background, a hand with the wrong number of knuckles, jewellery that changed shape between the thumbnail and the crop. We shot the campaign the following week.

That conversation now happens almost monthly. Every brand I work with in Mumbai is being told by someone that AI will make photographers obsolete, and every one of them is quietly nervous about it. I'm not. But I'd be lying if I said nothing had changed. The tools have already rewritten parts of my workflow, killed some line items on my invoices, and created others. The honest answer to "does AI replace the advertising photographer" is neither the panic nor the dismissal — it's a careful map of what genuinely shifts and what stubbornly refuses to.

This is that map, written from inside a working commercial practice rather than from a conference stage. I've spent the last year folding these tools into real client jobs, and what follows is what I've actually learned about where the machine helps, where it embarrasses you, and why the role of the advertising photographer Mumbai brands trust is changing shape rather than disappearing.

What AI genuinely changes in the workflow

The first place AI showed up in my studio wasn't image generation at all — it was post-production. Retouching that used to eat a full day now takes a couple of hours. Generative fill cleans up a stray cable or a crumpled backdrop seam in seconds. Masking, which I once did by hand on every hero frame, is now a single click that's accurate enough to need only minor cleanup. Sky replacement, background extension for a banner crop, removing a reflection of the softbox in a watch face — these are no longer the painful parts of the job.

This matters more than it sounds. A large chunk of a commercial photographer's cost was never the shutter press; it was the hours afterward. When I quote a campaign for an FMCG brand like a Mamaearth or a Bombay Shaving Company, the retouching estimate used to be a third of the budget. AI has compressed that meaningfully, which means I can either pass savings to the client or spend the reclaimed time on the things that actually differentiate the work.

The second shift is pre-production and ideation. I now build mood boards and lighting tests in a generation tool before anyone steps on set. If a client briefs a "warm, golden-hour, candid" mood, I can generate ten interpretations of that phrase in an hour and we argue over real pictures instead of adjectives. It doesn't produce the final image — it aligns the room. The brief stops being a misunderstanding waiting to happen.

The third, and the one most people fixate on, is full synthetic generation: invented models, invented products, invented worlds. This is real and improving fast. For certain jobs — a conceptual background, a surreal campaign key visual, a texture or sky I'd otherwise license — it's genuinely useful. But it's also where the embarrassing failures live, and where the gap between a phone-screen thumbnail and a printed 6-sheet is widest.

"It is the photographer, not the camera, that is the instrument." — Eve Arnold

The quote is older than any algorithm, but it has never been more accurate. The tools changed; the instrument — judgment, taste, the decision about where to point attention — did not.

What AI cannot touch

Here is what a year of real jobs taught me the machine cannot do. It cannot stand in a room with a nervous founder and a sceptical brand manager and read the temperature. It cannot notice that the model's shoulders tense every time you mention the deadline, and loosen the moment you make her laugh — and that the loose frame is the one that sells. The single most valuable thing I do on a shoot day is human direction, and there is no prompt for the trust that produces a real expression.

It cannot handle the physical truth of a product. A founder hands me their actual serum bottle, with its actual glass, its actual label registration, its actual cap that has to look exactly like the one a customer will unbox. AI will happily invent a bottle that is "like" theirs — close, plausible, and legally and commercially useless. Product photography for a real D2C launch is about fidelity to a specific object, and fidelity is precisely what generative models are worst at.

It cannot give you authorship you can defend. The legal ground under AI-generated images is still shifting, and a brand that builds a campaign on purely synthetic imagery is taking on usage and ownership risk it often doesn't understand. When I license images to a client, they know exactly what they're buying and that it's clean. I wrote more about how that side works in my piece on photo licensing and usage rights, because it's the part of the business that AI has quietly made more important, not less.

And it cannot do Mumbai. It cannot put a model on a Colaba rooftop at the exact minute the light goes from harsh to honeyed, with the real texture of a real city behind her. The specificity of place — Bandra's painted walls, the industrial bones of Lower Parel, the grey churn of Marine Drive in monsoon — is something clients increasingly want precisely because it can't be faked. As a commercial photography Mumbai practice, location authenticity has become a selling point rather than a logistics headache.

"To photograph is to hold one's breath when all faculties converge to capture fleeting reality. It's at that moment that mastering an image becomes a great physical and intellectual joy." — Henri Cartier-Bresson

That convergence of faculties in a single fleeting moment is the thing being sold. A model that averages a billion images cannot hold its breath.

A clear-eyed comparison

Clients ask me to lay this out plainly, so here is the comparison I actually use in pitch meetings.

The table isn't an argument that AI is bad. It's an argument that the two things are good at different jobs. The brands that get this right — the Nykaas and the Sugar Cosmetics of the world, who run enormous volumes of imagery — increasingly use synthetic tools for low-stakes social filler and book real shoots for the assets that carry the brand. The mistake is treating them as substitutes rather than as a portfolio.

How the role of the photographer is shifting

If the camera work and retouching are partly automatable, what is the advertising photographer Mumbai brands keep hiring actually for? The answer is that the centre of gravity has moved from execution to judgment.

My value used to be heavily weighted toward technical craft — knowing how to wrap a hard light into a soft one, how to expose for a difficult skin tone, how to get a glossy product clean of reflections. Those skills still matter enormously, but they're increasingly table stakes. What clients now pay a premium for is the part that sits above the tools: the ability to take a vague commercial objective and turn it into a specific, ownable visual idea, then execute it with enough control that it works across a billboard, a bottle and a feed.

In practice this means I spend more time as a creative director and less as a button-pusher. I'm choosing which frames to generate as references, then which to shoot for real, then which to retouch with AI assistance and which to leave honest. The throughline is taste — and taste is the one thing the tools cannot supply, because by design they regress toward the average of everything they've seen. Average is the enemy of advertising.

The fashion side has shifted similarly. AI can generate a "fashion" image, but it cannot build the relationship with a stylist, a designer and a model that produces a coherent fashion photographer Mumbai campaign with a real point of view. International houses learned this first — the reason a Calvin Klein or a COS image reads as that brand is intentional human authorship, not a prompt. Indian fashion and beauty brands are arriving at the same conclusion, which is why I've seen demand for considered, art-directed shoots rise rather than fall over the past year.

How I actually use AI on a job today

Concretely, here's the workflow on a typical campaign now. During pre-production I generate mood and lighting references to lock the creative direction with the client — fast, cheap alignment. On the shoot day, nothing changes: real product, real talent, real light, real direction, because that's where the value the client is paying for actually gets made. In post, I lean on AI for the grunt work — masking, clean-up, generative fill on backgrounds, occasionally extending a frame for a crop ratio I didn't shoot wide enough for. What I never do is let a synthetic detail stand in for a real product feature or a real human expression on a hero asset.

The result is that I deliver faster and the client's money concentrates on the irreplaceable parts. That's not AI replacing the photographer. That's AI removing the drudgery so the photographer can spend more of the budget on craft. Brands like Boat and FabIndia don't want fewer real pictures — they want their real pictures to land harder, and the reclaimed time goes exactly there.

The photographers who will struggle are the ones whose entire offer was execution — clean, competent, undifferentiated images that AI can now approximate. The ones who will thrive are the ones who were always selling judgment, relationships and a point of view, and merely used the camera to deliver it. If that sounds like a hard line, it's the same line every technology shift has drawn: it doesn't replace the craftsperson, it replaces the part of the craft that had become mechanical.

FAQ

Will AI replace advertising photographers in Mumbai?

No — but it's reshaping the job. AI is replacing the mechanical parts of the workflow (some retouching, masking, reference-building) and handling low-stakes filler imagery. It cannot direct real talent, render a real product with the fidelity a launch needs, or deliver legally clean, brand-specific hero assets. The photographers being squeezed are those who only sold execution; those who sell judgment and authorship are seeing demand rise.

Can I just use AI images for my product campaign instead of a shoot?

For mood boards, conceptual backgrounds and some social filler, yes. For your actual product, no — AI invents details and won't match your real bottle, label or finish, and it fails on inspection at large print sizes. Most brands now use a mix: AI for ideation and low-stakes content, real shoots for the assets that carry the brand.

Does using AI in post-production make the photos "not real"?

Not in any way that matters commercially. Retouching has always been part of advertising photography — AI just makes the routine parts faster. The line I hold is that I never let a synthetic detail stand in for a real product feature or a real human expression on a hero frame. The image is built on a real shoot; the tools speed up the cleanup.

Is it cheaper to hire a photographer who uses AI tools?

Often, yes — because AI compresses the retouching hours that used to be a large share of the budget. That saving either lowers the quote or gets reinvested into more craft on set. Either way the client gets more value per rupee, which is why I've folded these tools in rather than resisting them.

The campaigns that will age well over the next few years won't be the ones that used AI the most or rejected it entirely. They'll be the ones that knew exactly where the machine helps and where only a human standing in a room with real light and real people can do the job. That judgment is the whole craft now — and it's the part worth hiring for. If you're planning a campaign and want to talk through where real photography earns its place, take a look at the recent work or get in touch.

Harnesh Joshi is an advertising and fashion photographer based in Lower Parel, Mumbai. View the portfolio or get in touch to discuss your next campaign.