It’s not just you, images on the internet look weird now

You’ll have seen it by now; the invasion of the AI stock image. Major stock photo suppliers are gradually introducing these uncanny-valley creations into their libraries. But while they are undoubtedly cost-effective, the customer feedback may be less positive.

In a recent report by Getty Images, people do not like being “lied to” when it comes to AI photos. 87% of respondents say it is important for an image to be authentic.

Whether they use it for marketing, website design, or creating content, most online companies rely heavily on stock imagery to sell their products and enhance their services.

For this group, the question of whether to trust AI-generated visuals is becoming critical.

The AI image conundrum

AI-generated images are a shortcut for SMEs. Small firms cannot afford expensive photoshoots with human models. AI shots are often more affordable than human stock photography, and can be created in the fraction of a second with a simple typed prompt.

Yet, as some have discovered the hard way, they are also risky. Queensland Symphony Orchestra made headlines when it used an AI-generated image described as “two people having a date at an indoor classical music romantic concert” for a Facebook advert.

Classical music lovers in north-east Australia were bemused to scroll past the image of two plasticine-looking people with six fingers, listening to an army of identical violinists that inexplicably sat behind them in the stalls.

Image credit: facebook.com/QSOrchestra/

Other famous offenders include the confectionary car crash of Glasgow’s ‘Willy Wonka’ Black Mirror-esque experience, which was advertised using AI-generated gibberish.

Getty and Shutterstock AI

The above are particularly egregious examples of bad AI stock. But the controversies may become common given the many image providers leaning on AI to pad out their offerings.

They include Getty Images, which has partnered with Picsart to build its own custom model. Using stock photos from Getty’s archives, the collaboration is aimed at giving SMEs affordable access to promotional and supportive materials for their online content.

It’s not just Getty. Shutterstock last year unveiled AI tools that will allow businesses to generate alternate versions of real photos, as well as expand the background of a picture.

Crucially, the images used are licensed with coverage provided by the platform. That means they are free for commercial use and users cannot be in breach of copyright laws.

However, firms may not be safe from audience backlash. Queensland Orchestra received huge backlash for its use of the image, which was sourced from Shutterstock, which many saw as out of tune with the organisation’s mission statement to support the arts.

“Next time pay photographers,” one comment read, while another said “terrible – literally an arts organisation not using artists.”

Time to take stock of AI

Almost 90% of consumers globally want to know whether an image has been created using AI, according to the Getty Images survey.

Firms that want to take advantage of cost-effective AI must be cautious about how a rapid rollout might impact customer needs and expectations.

The images may be fake, but the struggle is real for today’s businesses. The question of whether to adopt AI, or wait and see if the bubble will burst, haunts many leaders.

It’s not just imagery. HMRC recently found itself in hot water after it tried to terminate its support helpline earlier this year and replace it with an automated service.

As the digital world becomes saturated with AI content and services, businesses must tread carefully. The question of whether to adopt AI is no longer a matter of if, but how. Misguided use of AI could prove as costly as it is embarrassing.

Ultimately, the customer must come first. By putting audiences at the heart of their strategy, companies can harness the technology’s potential while safeguarding their reputation.