5 Signs to Identify AI Content on Instagram
When you want to find out an answer quickly, which do you go to first – Google or Chat GPT? Gemini?
Artificial Intelligence is integrating itself into every aspect of life and speeding up processes that used to be time-consuming.
The marketing space has been overtaken with AI as many are using it for project management tools, scheduling, and the mundane tasks that come with doing admin. A huge area that AI has slid into is the content creator space, especially on Instagram.
Meta has unlocked many features this year, from chatbots to AI-Generated Images.
AI is helping businesses and creators create content in a quicker timeframe but what are the concerns around AI? How do we know what’s real and what’s AI-Generated?
In this month’s blog, we’re going to discuss the advancement of AI, 5 signs that an Instagram post was made using AI, and what does “made with AI” mean attached to an image.
Let’s dive into it.
The Advancement of AI on Instagram
Did you know AI integration on Instagram began in 2010?
The filters to adjust your images brightness or add a different tone to the image is all done by AI.
The suggested accounts that pop up in your feed exist because of Instagram’s machine-learning algorithm.
When you hear people talk about keywords, SEO, hashtags, this is the reason. In 2010, the algorithm was at its most basic, using image recognition to categorize photos based on content.
As Instagram switched from its basic capabilities to more advanced, machine learning techniques like “word embedding” analyze the content and categorize it in terms of relatedness.
Who said words don’t matter? Your bio (the small description you have on your instagram), the captions, the hooks, the copy under the post is all being scanned to categorize your account.
So pay attention to your words!!
The transformation of AI capabilities since 2010 to the present moment in 2024 have become far more advanced at a rapid pace.
Let’s take a look at the timeline.
Between 2010 and 2014, AI was focused mainly on image enhancement like brightness and contrast, cropping your photos, personalized content based on your interactions and preferences.
From 2015 to 2019, the features began to roll out quicker.
- The explore tab was enhanced using a more advanced machine learning algorithm, improving personalization and content discovery.
- “Word embedding”, keywords and hashtags, helped analyze the content to recommend accounts based on interactions.
- Seed accounts, created by instagram,used to test new features, algorithms, or functionalities before rolling them out to the general public.
- AR and Virtual reality began to be rolled out as well as the use of NLP, Natural Language Processing, machines to scan for offensive comments and limit spam and misinformation.
- The birth of IGTV with AI recommendations personalized video content for each user based on interactions and preferences.
- AI-Enhanced shopping features were created to recommend products and target ads based on user behavior.
From 2020 onwards, the creation of tools for content creation like automated caption suggestions and hashtag recommendations began to roll out as well as conversational AI assistants that would answer queries, provide content recommendations and help in generating posts.
In 2024, the rollout of Meta AI on Instagram is even more powerful compared to what it was in 2010.
The implementation of AI on the platform has saved content creators, businesses, brands hours of work but a major issue that has caused controversy is when people use it for potential harmful reasons.
Because of the advancement of AI, it now has the ability to create images that look eerily real, making it harder to know what’s AI created and what’s REAL content.
This is down to generative AI which works by creating new data that mimics human creations.
If you struggle to distinguish between superficiality and reality, never fear because here are our top five signs that an image was AI-Generated.
The 5 signs of an AI-Generated Image
We all want to know whether an Instagram post is made with AI because Instagram is known to be the “highlight” reel, the best bits.
Ensuring the user is informed about what’s AI-Generated and what’s not is something Mark Zuckerberg, head of Meta, wants to make sure is maintained.
Here’s five pointers to spot AI going from the easiest to hardest to spot.
1. Distorted Backgrounds:
If a door looks like it has a bump in it or is twisted into another shape, it’s a clear indication something was done to the image. This is one of the easiest ways to spot an image was
2. Something feels.. Off
We wrote about AI influencers in our last blog (read more on it here) and there’s a slight unease when looking at the image.
They say you can read a person by looking into their eyes. When you see an AI-Generated image of a “person”, there’s nothing.
Almost like a blank stare. Like it’s not… real. AI influencers on Instagram have risen in popularity with many brands like Chanel and Red Bull using them for online promotions.
3. Pay attention to the small details
We’re scrolling 24/7 (hopefully not!), and are scanning images and text at a quicker pace. This means we miss the finer details.
Don’t take an image at face value if you feel like something isn’t right. AI images are created by taking data from other photos and creating a new image.
AI programs often struggle with details you may see less frequently in other pictures.
AI tends to get the smaller details like logos and text wrong as you can see below.
4. If it looks too good to be true, it probably is
As much as we want to look perfect all the time, we don’t. We’re humans, imperfection is perfection.
You can tell if something is too good to be true if the image looks smooth and no texture because well, we have texture.
So don’t be fooled by an image looking picture perfect because it most likely isn’t real.
5. Reverse Search
You know the TV show “Catfish”? If you don’t, it’s a programme where the hosts bring on a person who writes to the show about their relationship.
They write in because they’ve never met the person and want to make sure they are the real deal. One of the first things the hosts do is reverse search the image.
They do this because they want to see if the person is using someone else’s photos.
This is the same when it comes to images on Instagram.
If in doubt, reverse image search because if it is AI-Generated, there won’t be a lot of information which tells us something is fishy.
“Made with AI”
You might have spotted on Instagram recently under some images and videos are a tag saying “Made with AI”.
This feature was recently introduced however, there has been some controversy especially for photographers.
How it works
When you upload an image to Instagram, the algorithm will read the metadata of the image to clarify whether it’s real or AI-generated.
This is great BUT what if you just enhanced the brightness? Enhanced the sound so it’s clearer? Removed the blemish from your face that was annoying you?
Instagram smacked the tag of “made with AI” on it. The problem for photographers is the wording.
Saying an image was made with AI sounds like it was superficial, that they didn’t put the hard work into capturing a beautiful moment like a wedding or someone’s birthday.
Some argued for it to be changed to “edited with AI” which clarifies that the image has been enhanced but it was taken professionally, not created superficially.
Because of the backlash, from July this year, the made with AI feature has been changed to AI Info.
This is a step in the right direction to distinguish the difference between an image being AI-Generated and one that’s a real-life photo.
GoViral Conclusion
Our advice is to always be cautious about images that are posted online.
If something isn’t sitting right with you, there’s a reason why. Our five pointers will give you more clarity around AI-Generated photos and with the introduction of “AI Info” on images,
It can help you distinguish between superficial and reality.
As many new features are being added to social media platforms every day, it’s important to stay up-to-date with the new trends and implementations.
If you want monthly updates on the latest trends in social media and AI, subscribe to our monthly newsletter.
Will Meta's AI Features Impact the Future of Influencer Marketing?
In the last few years, Meta AI has introduced different features, disrupting influencer marketing and engagement between creators and fans on Instagram. Between the introduction of virtual humans and the Creator AI studio where creators can train AI bots to respond back to fans, the line between reality and virtual reality becomes blurred.
AI has transformed our lives, for the better and for worse. Tasks are completed quicker and more efficiently, questions are answered by the speed of light (well, kind of) so no more rummaging through papers to find an answer, and transcriptions of meetings helps us remember what was said in the meeting at 9am on a Monday morning.
The creation of AI has helped us more quicker and faster than generations before but are we taking it a step too far by creating virtual humans and training AI bots to be like us?
In our blog, we’ll talk about how influencer marketing is changing for both brands and influencers and how the new features introduced by Meta AI are changing social media forever.
Traditional Advertising is out, Influencer Marketing is in
When we think of the word influencer, we think of the modern day person who has amassed a large following on social media platforms through niche-specific content that targets a specific audience.
But if we think of the word “influence”, it can be used for anyone that’s able to influence a large amount of people and get them to do something.
The influencers in our life before social media were the Church, the government, even our parents but fast forward to the present day, creators on the internet now have a bigger influence on us than institutions and governments.
The mid-2000’s was when Facebook and Twitter were born and the term “influencers” started to gain traction.
As people started to grow their followings and see quick engagement, brands noticed a trend appearing that led to them partnering with influencers to expand their reach and increase sales.
Influencer marketing has become more popular than traditional marketing because of the deeper connection made through social media, compared to billboards and AD’s on TV.
People trust other people’s opinions before buying. How many times have you checked reviews before you bought something?
Some influencers have come under fire for collaborating with brands that don’t fit their niche, giving the impression that they’ve partnered for the money.
The importance of choosing the right influencer for the right campaign can make or break the trust between the influencer and their fans.
In the last year, money transparency around how much influencers are making has been exposed with some influencers that have 10 million followers are asking for 10,000 USD for just one video. Although it comes with recognition and visibility, this doesn’t guarantee sales.
The introduction of AI-Generated Influencers have risen in the past few years making it more lucrative for businesses to cut down on costs and logistics but there are ethical concerns to consider.
The Rise of AI-Generated Influencers
If you search “Top 15 AI influencers” on Google, a list of AI-Generated influencers will pop up on your screen.
If you check out those influencers, they aren’t real people but a virtual human that has built a large following.
Brands like Chanel, Red Bull and Calvin Klein have used these AI-Generated influencers for online promotions.
One popular influencer is Zara Shatavari who has 19.7k followers on instagram and is India’s first AI generated brand ambassador.
Zara was created to spread awareness for those living with polycystic ovary syndrome (PCOS) and runs a health-related blog, posts about travel and fitness.
Rahul Choudry, who created Zara, wanted to partner with a specific celebrity with a hormonal imbalance but chose to create an AI-version that was much cheaper, always available, and gives him the ability to adapt it when needed.
Another AI influencer goes by Reembot on Instagram focusing on fashion and lifestyle content.
Influencers post their day in their lives, their workout routine, their lifestyle and fans get invested because of the relatedness of it but how can an AI influencer that doesn’t live in the real world create that same relatedness when we know it’s not real?
Backlash from people online have called out brands for choosing to invest in virtual humans, rather than real influencers.
As more brands continue to choose AI-versions to cut down on costs and logistics, is REAL influencer marketing dead?
The Future of Engagement
Along with AI Influencers, Meta’s CEO, Mark Zuckerberg, rolled out creator AI for select creators in the US where they can create AI bots of themselves.
He’s testing the feature out to see how people interact with it first before he rolls it out to other countries.
The feature will allow creators to train an AI version of themselves to respond back to comments and direct messages that are sent from fans.
Creators will train the AI bot to replicate their personality and personalize their messages to maintain authenticity.
Creators receive a huge amount of messages every day, sometimes asking for something simple like a link or a frequently asked question.
Meta is committed to enhancing engagement while also being open about the fact that the fan will be interacting with an AI bot.
This is extremely important that the fan KNOWS it’s not the creator himself responding but that the responses are personable and are scripted out to respond in this way.
However, the complication comes when a fan is asking for a more complex question that requires a deeper answer that’s subjective to the influencer.
This feature is still in BETA so it’ll be interesting to see how it plays out and how creators can maintain the level of authenticity and engagement needed.
GoViral Conclusion
AI has integrated itself into our lives where reality and virtual reality are becoming nearly indistinguishable. Meta AI continues to advance and new features are emerging quicker than before.
Social media has changed drastically since 2007 at a quick speed and doesn’t seem to be slowing down as Mark Zuckerberg has big plans for the Metaverse with talks about holographic glasses projecting real people in virtual reality (VR).
The opportunities and possibilities that come with this new technology is immense and it feels like it’s just the tip of an iceberg.
However, the question still stands. How far is too far?
Subscribe to our monthly newsletter to receive social media upates, trends, and marketing insights.