I’ll start with a confession: I use AI, and I’ve been using it for years! AI wasn’t invented by ChatGPT. It’s been around for decades, serving as a fundamental component of many tools we use daily and the social media platforms themselves. So, anyone who tells you they haven’t or don’t use AI gives a sure sign that they don’t know what AI is!
I’ll go further with my confession. My grammar isn’t great. Every school report highlighted that my storytelling was excellent, but my grammar was poor. And with spending more time representing my county and country at football, the grammar situation never improved. So I, for one, have been using Grammarly for many years. It’s a vital tool in helping me craft long-form content that’s presentable. I’ve also been known to check social posts where I have a doubt over my use of grammar.
My Definition of Using AI
Let’s establish something fundamental: The originator of the idea, theme, or story is the creator and therefore owner of the IP (intellectual property). The fact that you use tools to help you research, fact-check, and make sense of your thoughts isn’t bad. It’s fundamental in helping you communicate in the most effective way.
Think about it – we’ve never questioned whether using spell-check diminishes authorship. We don’t debate whether using design software makes you less of a creative. So why the sudden panic about AI?
The Verdict Upfront: It’s Both Good and Bad
Is the use of AI in social media beneficial or detrimental? Let’s cut to the chase – the answer is both!
As the versions of AI that have suddenly “changed everything” (according to some) are Large Language Models, then in order to know if it’s good or bad, you first need to know what content is good or bad! Therefore, we’re back to the original creator – does that person understand the topic, subject, or story they’re trying to tell well enough to know whether ChatGPT, Claude, Perplexity, etc., helps craft what they’re trying to communicate?
The Bad: A Cautionary Tale from a University
Let me share an example that perfectly illustrates the bad. For a university client, I frequently received campaign and subject area marketing plans from one team member that had been directly lifted from ChatGPT. Complete with American spellings using z’s and all!
If this person were a student, they would have been dismissed from their course. I believe this particular university was one of the first in the UK to do just that for AI plagiarism. And yet, here was a senior member of the admissions team churning out work that clearly came from a one-line prompt: “Write me a marketing plan for [subject].”
It was incorrect in so many places, highly irrelevant, and had no use whatsoever! This, you must agree, is bad! It proved a clear lack of knowledge by the originator. They couldn’t evaluate whether the output was accurate because they didn’t understand the subject matter themselves. They couldn’t spot the glaring errors, the American context in British education, or the generic content that meant nothing to their specific institution.
This is AI at its worst! A crutch for incompetence rather than a tool for enhancement.
The Good: AI as a Collaborative Partner
Now, had this university marketing person gone through proper steps and used the platform as a tool to support their work, I might have been more forgiving. When you understand your subject and use AI thoughtfully, it becomes transformative.
Here are five best-practice steps for getting output from AI that delivers the most human-oriented work whilst remaining on point for your subject area:
1. Start with Expertise, Not Ignorance
Before you prompt any AI, you must understand your subject. AI should amplify your knowledge, not replace it. If you can’t evaluate whether the output is accurate, relevant, and appropriate, you’re not ready to use AI for that topic. Do your homework first.
2. Provide Rich Context and Constraints
Don’t use lazy one-line prompts. Feed the AI comprehensive background: your audience demographics, brand voice, cultural context (specify UK English!), industry-specific terminology, and current market conditions. The more context you provide, the more relevant the output. Think of it as briefing a talented but unfamiliar colleague.
3. Iterate and Refine Ruthlessly
Never accept the first output as final. Use AI’s initial response as a starting point, then refine through multiple iterations. Challenge it, correct it, push it further. Ask for alternatives, request different angles, demand evidence for claims. This iterative process mirrors how we develop ideas with human colleagues.
4. Inject Your Unique Perspective
After AI provides the framework, infuse it with your personal insights, real examples, and authentic voice. Add the stories only you can tell, the observations from your specific market, the nuances that make content genuinely valuable. AI can’t replicate your lived experience; that’s your superpower.
5. Fact-Check and Localise Everything
AI often defaults to American spellings, references, and contexts. Always review for cultural relevance, verify every statistic, check every claim, and ensure examples resonate with your specific audience. For UK content, this means British spellings, UK regulations, local case studies, and cultural references that actually land.
The Nuanced Reality
The truth is, most of us operate somewhere between these extremes. We’re neither completely replacing our thinking with AI nor ignoring these powerful tools entirely. The key differentiator? Knowledge and intention.
When a social media manager uses AI to generate twenty headline variations and then selects and refines the best based on their understanding of their audience, that’s good use. When someone copies and pastes AI output without understanding whether it’s even accurate, that’s bad use.
Moving Forward: Ownership and Authenticity
Remember my definition: you own the IP when you’re the originator of the idea. AI doesn’t diminish that ownership when used properly. It enhances your ability to express and refine your ideas. But you must remain the architect, the editor, the quality controller.
The proliferation of AI in social media isn’t going away. The question isn’t whether to use these tools, but how to use them responsibly. Those who understand their subjects deeply will use AI to communicate more effectively. Those who don’t will be exposed by their inability to spot AI’s mistakes and limitations.
The Bottom Line: We’re at a Crossroads
Right now, LinkedIn is drowning in identical “I’m humbled to announce” posts that could have been written by the same bot. Instagram captions are becoming eerily similar, all following the same formulaic patterns ChatGPT loves to produce.
We’re at a crossroads. Either we use AI to elevate human creativity and help more authentic voices be heard clearly, or we let it turn social media into an echo chamber of mediocrity.
AI in social media is both good and bad because people are both competent and incompetent, thoughtful and lazy, innovative and derivative. The tool simply amplifies what’s already there.
My grammar may never be perfect, but with AI’s help, my ideas, my original ideas, can reach you more clearly. That university person’s grammar was perfect too, but their content was worthless because they contributed nothing of themselves to it.
The difference? I know my subject. I know my audience. I know when AI is helping and when it’s not. And most importantly, I know that the story, the insight, the value, must always come from me.
The choice we make now will determine whether social media remains a space for genuine human connection or becomes a wasteland of AI talking to AI whilst humans scroll past, desperately searching for something real.
Choose wisely. Your audience is counting on it.