AI in Focus: Navigating the Dual Role of Ally and Adversary in Photography’s 2024 Landscape

When you purchase through links in our articles, Future and its syndication affiliates may receive a commission.

After yet another year of rapid advancements in AI, we’ve arrived at the moment in the year where we conventionally halt and inquire: what just took place? Merely two years ago we discussed how AI was transforming photography with the surge of AI image creators. Then the prior year, we observed AI becoming mainstream and beginning to transform photographers’ processes.

In 2024, these advancements persisted. AI image generators improved and additional AI tools were incorporated into photo editing applications. We noticed some initiatives to introduce transparency for the emerging medium but also prospective hurdles for photographers.

Here, I will summarize the key advancements in AI within photography during 2024 and reflect on where things could head next…

That appears to be AI… or does it?

By the conclusion of 2023, some of the finest AI image generators had already produced images that could occasionally be difficult to differentiate from photographs. Advancements took another significant leap in 2024.

With Black Forest Labs’ Flux, individuals have been generating exceptionally realistic outputs made feasible through the application of fine-tuning scripts known as LoRAs (Low-Rank Adaptations) – extra custom weights that can deliver added detail. Enthusiasts who experimented with open-source Flux on their private computers began creating their own LoRAs to achieve particular styles – for instance, hyperrealistic portraits (refer below).

There remain signs such as nonsensical text and occasionally bizarre anatomy or lighting, but we are approaching an era where the camera is no longer the exclusive means to generate an image that resembles reality.

Progress will undoubtedly continue in 2025. Curiously, some developers keep referring to ‘photoreal’ as the ultimate target to strive for. Google showcased its Imagen 3 by illustrating its capability to create vintage-style “Polaroids”. However, there’s also an impression that AI has nearly perfected still images and that video may be the next challenge.

Google is evaluating an upgrade for its Veo generative video platform that enables the software to “grasp the unique language of cinematography” and replicate specific lenses. Meanwhile, Midjourney harbors even grander aspirations, intending to expand into the generation of 3D VR environments.

Everyone’s a photo editor

If an image originated as a photograph, but the photographer inserted a completely different sky using AI, added a subject via AI, and eliminated a few trees with AI, how much of the image remains a photograph? Compositing has always been part of digital art, but in 2024 it reached people’s holiday memories thanks to the surge of AI editing frameworks.

AI applications can now be found in the majority of the premier photo editing software. Photoshop has incorporated comprehensive AI image creation, alongside the pre-existing Generative Fill and Expand features, placing text-to-image generative AI into the hands of designers and brands everywhere.

Concurrently, AI has democratized basic retouching for anyone possessing a smartphone through manufacturers’ camera applications, most of which now include at least basic AI object and background elimination.

As a consequence, photo editing is no longer perceived as sorcery, and this shift carries implications. Photographers report clients attempting to reduce costs by claiming they will handle the editing themselves, with some even returning professionally edited photographs to the photographer with their own alterations. Ouch!

An AI photo-editing solution for every need

In addition to ‘general’ AI image generators, an extensive range of specialized AI editing tools exists for more precise applications. There are modular software like the Retouch4me apps for retouching, along with Topaz’s noise reduction and sharpening utilities.

There are also solutions like renderspace.co that allow the virtual staging of interior shots with furniture, an overwhelming number of applications for creating AI headshots for CVs, and offerings like Pincel that can showcase a clothing brand’s garments on an AI model, potentially undermining opportunities for both photographers and models (apologies if this casts a shadow over your celebrations).

Photos that animate… yet aren’t videos

AI is also shaping trends in social media. You’ve likely noticed those AI ‘videos’ appearing in your feeds. AI video generators such as Runway and Sora are merging the boundaries between photography and video (often with ‘blur’ being the key term), by introducing motion into static images.

The outcomes can be quite alarming, yet the technology is set to improve, which could reinforce a shift in preference towards moving images over still photographs on social platforms while also potentially altering workflows for those in mixed media.

Brands recognize that AI is adequate

AI image generators have been readily accessible for a few years; however, major brands have predominantly kept their distance. Concerns over copyright and apprehension about adverse public responses were too pronounced. This began to change in 2024.

We have observed media organizations employing AI imagery to complement articles and brands utilizing AI-generated visuals for everything from social media posts to advertisements and websites. Even Coca-Cola reimagined its iconic Christmas advertisement using AI-generated video (so much for “the real deal”).

Nevertheless, AI imagery has not been entirely welcomed by the public. Companies as diverse as Lego, Netflix, Skechers, Transport for Ireland and the Tedeschi Trucks Band have faced backlash on social media for utilizing AI art, indicating that employing AI for final assets remains a perilous endeavor for brands.

Instagram’s AI tag frustrates all

In 2024, Instagram and Facebook made an insincere effort to mitigate the discourse surrounding AI content on social platforms by launching a ‘Created with AI’ tag. It promptly offended some photographers who believed their application of AI editing tools to be too minor to deserve such condemnation. Consequently, Instagram fumbled the situation even further, altering the terminology to “AI info,” igniting a discussion regarding how much AI constitutes an excess of AI.

Additional challenges exist. Meta does not automatically highlight AI content unless it has Content Credentials assigned, leaving it up to users to be truthful and tag it themselves. As if anyone is going to do that!

Perhaps in 2025, we might have to ponder whether the landscape should be reversed. A ‘Crafted by a human with a camera’ label could elevate conventional photography above AI-generated content.

Photographers seek alternatives

Instagram has established itself as the preferred social platform for a multitude of photographers, even birthing entirely new styles of photography alongside the emergence of the photographer-travel influencer. However, with the advent of AI – and AI influencers – now permeating feeds, it is becoming increasingly challenging for photographers to gain visibility.

Meta’s rules permitting it to utilize pictures from public accounts for training its AI image generator were the tipping point for some. The inquiry now is where to turn. Cara has surfaced as a favored option among creators due to its prohibition on AI content, restrictions on AI training, and the ability to incorporate Glaze (which provides misleading data to AI) into your creations. It’s accessible to photographers as well, but it lacks a broad following outside the realm of artists.

Explore our recent guide on what social network photographers should choose in 2025 for the comprehensive details.

Welcome the robotic photographer

With AI currently assisting in photo editing, some individuals believe it can also contribute to capturing images – and I refer not only to in-camera AI features like eye autofocus. Researchers created PhotoBot as a trial to eliminate the need for tripods and timers. Employing a visual language model, an object recognizer, and a large language model (LLM) to fetch reference images pertinent to a user’s request, it takes commands, analyzes the environment, and suggests a layout for a shot.

PhotoBot was merely a trial, yet it offered a glimpse into how AI might influence both photography hardware and software. Numerous other experiments are underway, including tools designed to automate photography for generating 3D visualizations – such as ActiveSplat, which navigates an indoor area to create a 3D depiction.

Event photographers who believe they occupy a segment impervious to AI’s advance can anticipate Tesla training Optimus to take photos.

What lies ahead?

Several of these advancements in AI might be discouraging for photographers, and it’s indeed accurate that AI and photography share a complicated rapport. However, it is also occasionally an advantageous one, enhancing creative opportunities and accelerating workflows.

Generative AI is undoubtedly exerting pressure on the demand for photography in certain spheres, or at least altering the market’s landscape to some degree, especially within the realm of generic stock imagery and lower-tier product photography. We can expect to witness an influx of niche AI products continuously intensifying this pressure.

Nonetheless, it’s not entirely doom and gloom. The reality that “it appears AI-generated” has become the ultimate criticism on social media (regardless of whether the image in question is AI or just overly edited) illustrates that, for many, authentic photography and genuine subjects maintain their value.

Photographers can distinguish themselves from AI outputs by cultivating a personal brand and narrative surrounding their creations while still leveraging AI to streamline the process if they choose.

You might also enjoy…

Desire to escape the digital realm entirely? Discover the best film cameras (and the finest 35mm film to accompany them) and the top instant cameras.