Most of the cropping, formatting and optimizing of photos and illustrations for website or social media can be taken over by AI today. Not everything, and not without prior checks on brand guidelines and image rights, but this is one of the tasks within content marketing and graphic work where the shift has already progressed far.
Three axes are decisive: structure, volume and creativity.
The task is highly structured. An image needs to get a certain ratio, a certain size, a certain file size. These are rules that are fixed in advance, not judgments that are made afresh case by case. The volume is high: a company that posts content weekly across multiple channels has tens to hundreds of images per month, each requiring the same kind of editing. Work that occurs frequently and follows fixed rules is precisely the work where automation delivers the most.
The creativity required is limited. Cropping and optimizing rarely involves designing something new; it is about fitting an existing image within existing frameworks. That is a different task from designing an image from scratch, which does require coming up with an idea, a composition or a style choice. That boundary runs through the entire field: taking over a format is something different from devising it.
Customer contact, physical actions and compliance score favorably, meaning there is little standing in the way. Error costs also score favorably: an incorrectly cropped image can usually be fixed quickly, without irreparable damage. Judgment scope scores average: with a standard photo for a product page there is little to weigh, but with an image touching on a sensitive topic or where brand identity carries significant weight, a judgment call about tone and context is still needed.
That is also where the difference between companies arises. A webshop with a fixed product photo style and a tight template for social media has hardly anything to weigh: everything is already determined, and AI works within those frameworks. An agency that edits varying image styles for a range of clients, with differing brand guidelines and sensitivity around brand identity, retains more moments where a human assesses the outcome before it is published. The task then remains largely automatable, but with a review step in between more often.
Take an example: an organization posts a product photo on the website daily and an adapted version on social media. In the past, an employee manually cropped both, adjusted the resolution and checked the format per channel. Today a system can carry out those steps once the brand guideline rules have been established: which ratio, which color profile, which minimum resolution. A human assesses beforehand whether the rules are set correctly and afterward, through spot checks, whether the outcome is correct. That is the second category within the shift: not fully automatic, not fully manual, but AI doing the work with human oversight that approves or rejects it.
This shift is connected to other work within the same set of tasks. Those taking over this task often also look at writing social media posts, because images and text arise within the same workflow. And because part of this work touches on published content, it is worth knowing how the GDPR relates to automating tasks, especially when images contain people.
The preconditions remain human work or stay in their proper place. Brand guidelines must be established before a system can work according to them: someone must determine what the brand identity requires, which colors and ratios are correct. Image rights must be checked; a system edits an image, it does not assess whether the image may be used. That is a legal question, not a technical one.
It is also not a reason to base personnel decisions on. This page describes what happens to a task, not what an organization should do with employees; separate legal requirements apply to that. And if an automated edit does go wrong, for example an image that ends up online incorrectly due to an error in the process, the question of liability when AI makes a mistake is relevant to know in advance.
Image editing rarely stands on its own. Companies looking at what AI can take over in the area of content often look, in the same movement, at analyzing website statistics or at how advertising budget is adjusted based on results. That is the common thread running through all similar questions: not one task is taken over, but a cluster of tasks within the same process, each with its own pace of shift.
Whether this task can indeed be largely taken over in your organization depends on how fixed your brand guidelines already are and how much variation there is in your image material. You can get an indication of this with the free quickscan: twelve questions, no account needed, resulting in an outcome showing which portion of the hours in your profile can be taken over by AI today. The full work scan, which breaks down the work of an entire company into tasks and calculates per task the impact on FTE capacity, is still under construction.
Vraag maar. Ik ken de kennisbank van deze site; wat ik niet weet, zeg ik erbij.
Answers come from this site’s knowledge base. Not tailored advice, and not a scan of your company.