Generative AI

AI Slop in Marketing Isn’t Failure: Ketepa, NTSA, and the Messy Middle

TLDR: AI-generated marketing can misfire — or create what’s called AI slop. Ketepa Tea, NTSA, and Coca-Cola show how public feedback and iteration turn AI missteps into stronger campaigns.

February is the month of love,  a season when brands lean heavily into emotion, gifting, and cultural moments. This year in Kenya, however, Valentine’s conversations went beyond flowers and chocolates. The Central Bank of Kenya cautioned against the misuse of banknotes amid the growing trend of money bouquets, highlighting how cultural moments, regulations, and public behavior intersect.

Against this backdrop, brands and public institutions turned to AI-generated images to participate in the moment. Ketepa Tea shared an AI-generated image of tea sachets arranged as a bouquet, prompting criticism around tone, timing, and authenticity. 

Around the same time, the National Transport and Safety Authority (NTSA) faced backlash after publishing an AI-generated #UsalamaBarabarani image that appeared to contradict its own road safety message. In both cases, audiences quickly identified the gaps and voiced their feedback.

These moments sparked a broader conversation about AI-generated images and videos in marketing and the growing phenomenon often referred to as AI slop. More than isolated missteps, they reveal how brands are learning to use AI in public — navigating feedback, iterating in real time, and refining their approach as audiences become more AI-aware.

What is AI slop in Marketing?

AI slop refers to low-quality, obviously AI-generated content that fails authenticity tests and lacks the human oversight necessary for quality marketing. It is characterized by visual tells like distorted hands, impossible physics, nonsensical text, and, perhaps most critically, missing cultural context that make content resonate.

AI slop is not simply “bad content.” It often results from:

• Over-reliance on AI tools
• Weak or rushed prompting
• Limited human oversight
• Speed prioritized over strategy

Crucially, AI slop reflects process gaps, not proof that AI itself does not work. Many of these outputs are not failures, but early-stage experiments published too quickly.

 But here’s what makes AI slop particularly problematic in 2025: today’s consumers are AI-literate. They can spot it instantly, and they expect better.

Why AI Slop Happens, Especially With Images and Video

AI slop is most visible in visual marketing for several reasons.

First, AI lacks lived context. Generative tools do not understand cultural nuance, regulation, emotion, or local realities. They work from patterns, not experience. Without human review, inconsistencies slip through.

Second, AI accelerates execution faster than judgment. What once took days of creative review can now be produced in minutes. When thinking does not keep pace with production, mistakes become public.

Third, visual content leaves no room for explanation. Audiences interpret images instantly. A single contradiction can undermine an entire message.

A Global Example: Coca-Cola’s Christmas Controversy and AI-Generated Creativity

In 2024 and 2025, Coca-Cola attempted to recreate its iconic “Holidays Are Coming” Christmas campaign using AI generation. The result? A soulless recreation that lost all the warmth and nostalgia of the beloved hand-crafted original. Visible AI artifacts and uncanny valley effects prompted hashtags like #BoycottCokeAI to trend. The message from consumers was clear: when you’re a heritage brand built on emotion, AI cost-cutting feels like betrayal.

Kenyan Case Study: NTSA and the #UsalamaBarabarani Campaign

Kenya’s National Transport and Safety Authority’s campaign  #Safewalking #UsalamaBarabarani (Road Safety) campaign, advocating for zebra crossing usage, offers a clear example of contextual misalignment. 

One of their AI-generated image shared on social media showed both a zebra crossing and a footbridge in the same frame. The public response was immediate: “Why would anyone use a ground-level crossing when there’s a safer footbridge option?” The image undermined the entire safety message, demonstrating AI’s inability to understand contextual logic. 

NTSA deleted the post, but the damage to credibility was done. Here is the next post they shared on X.

Every minute saved is not worth a life lost.
Use the footbridge, it’s there to protect you.
#SafeWalking #UsalamaBarabarani pic.twitter.com/NBsFRiLWW3

— NTSA KENYA (@ntsa_kenya) February 4, 2026

Ketepa Tea: Turning AI Backlash Into Iteration

Unlike the others, Ketepa’s response became a masterclass in crisis management. Ketepa Tea’s response stands out as a positive example of how brands can move forward after criticism.

After backlash over its AI-generated sachet bouquet, Ketepa did not defend the output or disengage. Instead, the brand listened. Here is the post that received a lot of criticism online.

Ketepa wishes to remind the public that section 1 of the Love Languages Code – aka the gifting of Ketepa Tea bags in decorative and romantic bouquet displays during this month of love is an enshrined freedom that all Kenyans across all walks can engage in without fear of… pic.twitter.com/c6rmS4ZGgb

— Kenya Tea Packers (@KetepaLtd) February 3, 2026

 

Kenya Tea Packers followed up with a new post featuring real flowers, real tea sachets, and a human presence, accompanied by the message “From AI to real tea.”

Ilibidi tutea.

Ketepa Tea Bouquet iko official,
romantic, inabrew love live.
Mapenzi ni vitendo… na tea ni lugha yetu.#KetepaTeaBouquet#ChaiNiKetepa#BrewLove pic.twitter.com/bbdehAiMOD

— Kenya Tea Packers (@KetepaLtd) February 6, 2026

This shift mattered. It showed responsiveness, humility, and an understanding that AI should support storytelling, not replace human connection.

The Messy Middle of AI in Marketing

These examples illustrate what can be described as the messy middle, the phase between experimentation and mastery.

Marketing has always evolved this way. Early social media content was awkward. Early mobile ads were intrusive. AI is no different. The difference today is visibility. AI compresses timelines and exposes learning curves publicly.

The case of Coca-Cola, Ketepa Tea, and NTSA, among others, who have boldly shared their AI-generated campaigns are example of brands that are no longer iterating behind closed doors. They are doing so in real time, with audiences actively participating in the feedback loop.

They proved that listening and adapting create more trust than getting it perfect the first time.

Each public failure trains marketers on what doesn’t work. Failed examples become case studies that refine our collective understanding of AI’s appropriate use. Community feedback improves prompt engineering industry-wide. The messy middle isn’t where brands get stuck; it’s where they build the institutional knowledge that leads to mastery.

Today’s Awakened Consumer

What makes 2026 different from even two or three years ago is audience sophistication. Some consumers can increasingly spot AI-generated content through visual tells, tonal genericness, and logical inconsistencies. More importantly, they expect better. The sentiment  you see on social media isn’t “don’t use AI”, it’s “if you’re using AI, at least make it good.”

This creates what is known as the authenticity paradox: brands must use AI for efficiency and creativity, yet audiences crave human connection and genuineness. The solution is not choosing between AI and authenticity. It needs brands to use AI as a tool while keeping humans as decision-makers.

Moving From AI Slop to AI-Smart Marketing

The difference between AI slop and AI success comes down to process. Successful AI marketing follows this framework:

• Generate multiple versions using AI for speed and scale. 
• Review through multi-stakeholder quality checks that include cultural consultants and brand guardians. Refine based on expert input and cultural context. 
• Test with small audience segments before full rollout. Approve only after final human sign-off confirming brand alignment. 
•  Monitor real-time sentiment post-publication. Iterate continuously based on learnings.

The critical element? Human oversight at every stage. AI should accelerate creativity, not replace judgment.

Brands must also build internal AI governance: documented learnings, prompt libraries with proven results, brand-specific guidelines, and team training on AI literacy. Quality thresholds should be established before generation begins, not after problems arise.

Conclusion

AI slop is not the end state of AI in marketing. It is a transitional phase that reflects experimentation, learning, and adjustment.

What defines strong brands is not the absence of mistakes, but how they respond. Brands that listen, iterate, and humanize their use of AI will build stronger trust and more resilient connections with their audiences.

In marketing, progress has never been perfectly polished. With AI, the learning just happens in public and when handled well, that visibility can become a strength rather than a weakness.

GPT-5 Explained: What Business Owners and Marketers Need to Know

OpenAI has just unveiled GPT-5, the latest and most advanced version of its AI model, and it’s already transforming the way we work, create, and communicate. Whether you are running a business, leading a marketing team, creating content, or simply trying to keep up with digital trends, GPT-5 promises faster, smarter, and more accurate results compared to previous models.

What Exactly is ChatGPT-5?

Think of GPT-5 as your new AI partner that’s sharper, more reliable, and more flexible than previous versions. According to OpenAI, GPT-5 is designed to handle complex tasks from drafting high-quality content and brainstorming creative campaigns, to analysing data and even interpreting images or charts.

Compared to GPT-4, GPT-5 has a stronger “long-memory,” meaning it can follow your train of thought over longer conversations, remember key details, and adapt its tone to match your style. It also does a much better job of reasoning through tricky problems, making it a powerful tool for marketers, business owners, and digital creators who want smart support on demand.

GPT-5 Access and Pricing Plans

OpenAI offers ChatGPT  through different subscription plans tailored to your usage needs:

• Free Plan: Provides limited access to GPT-5 with daily usage caps. Ideal if you want to try out the model casually. Once you hit your limit, you might get downgraded to a smaller, lighter version like GPT-5 Mini.

• ChatGPT Plus: This paid plan gives you higher usage limits and priority access to GPT-5. It’s perfect for marketers and creators who need reliable, faster responses throughout their day.

• ChatGPT Pro: Designed for power users and businesses, Pro users get the highest usage limits and access to GPT-5 Pro — a more capable variant built for handling complex, resource-intensive tasks with ease.
  
  

  ChatGPT pricing plans 2025

If you are a small business owner or a content team, these flexible plans let you pick the right balance of performance and cost for your needs. You can also explore the ChatGPT business plans

Key Features That Make GPT-5 Stand Out

Now let us look at some of the key features of ChatGPT-5:

1. Sharper Reasoning
   GPT-5 excels at multi-step thinking, breaking down complex questions and delivering more accurate answers.
2. Longer Context
   It can follow your conversation or project for much longer without “forgetting” earlier points, perfect for ongoing campaigns or content strategies.
3. Multimodal Power
   Beyond text, GPT-5 can understand and work with images, charts, and diagrams, giving you richer insights.
4. Smarter Adaptability
   It adjusts its style and tone automatically — whether you need professional copy for a client pitch or a playful caption for social media.
5. Better Accuracy
   Hallucinations (AI-generated wrong facts) are significantly reduced, making responses more trustworthy.
6. Custom Controls
   You can fine-tune how detailed or concise it should be, or how much “thinking” it puts into an answer.

The Limitations to Keep in Mind

While GPT-5 is impressive, it’s not perfect. It’s still not “true AI” that understands the world the way humans do and it can make occasional mistakes, especially with niche or tricky questions. This is evident with what some of the earlier users on Reddit are saying about ChatGPT-5.

Moreover, Free-tier users will also face daily limits, and while it’s much more accurate than GPT-4, it’s not immune to errors.

GPT-4 vs GPT-5: The Big Differences

Here’s how the new version stacks up side by side:

Feature	GPT-4	GPT-5
Reasoning Ability	Strong, but can struggle with complex multi-step problems	Much better at multi-step, complex reasoning
Accuracy	Good, but prone to occasional errors and hallucinations	Fewer errors, more factually reliable
Context Memory	Handles shorter conversations, limited long-term recall	Maintains context over much longer exchanges
Content Length	Up to ~100,000 tokens	Supports up to ~400,000 tokens or more
Pricing	Higher token costs, less cost-efficient	More cost-effective with flexible pricing plans
Adaptability	Can switch styles but may need prompts	Smooth and natural tone/style adaptation
Multimodal Support	Basic image and text handling (GPT-4o)	Advanced multimodal inputs (text, images, charts, audio, video)
Safety & Honesty	Occasional sycophancy and overconfident answers	More honest, transparent, and safer responses

Why This Matters for You

If you run a business, create content, or work in digital marketing, GPT-5 can save you time, sharpen your campaigns, and even spark ideas you might not have considered. At Samodigital Agency, we are already exploring ways to use GPT-5 for smarter automation, content creation, customer engagement, and market analysis.

The bottom line? GPT-5 is not just another tech upgrade. It is a productivity boost, a creativity partner, and a competitive edge rolled into one.

SEO in the Gemini Era: How Google’s AI Updates Impact Search Optimization

At the just concluded Google I/O 2024, Google announced expanded AI features for Google Search. How will this impact search engine optimization? Google has been making changes to Google search to improve how users find information in the search results.

The generative AI has even accelerated the improvement since the introduction of Gemini and the search lab’s search generative experience.

Here are some of the updates coming to Google Search in the Gemini Era

Google Search is introducing a new feature called AI Overviews.

AI overviews are summaries of search results that are designed to give users a quick overview of a topic without them having to click on multiple links. The feature has been tested in Search Labs, and users liked that they could get a quick summary and then links to learn more if they wanted. Google is rolling out AI Overviews to everyone in the US this week and expects to bring them to over a billion people by the end of the year.

Ability to Customize AI Overviews

Google Search will soon let you customize AI Overviews. You’ll be able to choose to simplify the language for easier understanding or get a more detailed breakdown of the topic

Search is getting smarter!

You can now ask complex questions in one go, thanks to a new AI model called Gemini. This means you can include all the details you want in your search, like location, criteria, and comparisons.

Google gives an example “For example, maybe you’re looking for a new yoga or pilates studio, and you want one that’s popular with locals, conveniently located for your commute, and also offers a discount for new members. Soon, with just one search, you’ll be able to ask something like “find the best yoga or pilates studios in Boston and show me details on their intro offers, and walking time from Beacon Hill.”

Plan Ahead!

Search can now help you create meal plans and trip itineraries. You can search for things like “easy 5-day meal plan for a group” and get recipe suggestions. Google says you can even customize the plan by asking for vegetarian options. You can then export the plan to Docs or Gmail. This feature is currently in testing but will be expanded to include other planning categories like parties and workouts.

Get AI-organized Results Pages

They’re introducing AI-organized results pages to help you brainstorm ideas. These pages will use AI to categorize different websites under unique headings, giving you a broader perspective on your search topic. This will be available first for searches related to dining and recipes, with plans to expand to movies, music, and more.

Google Search is getting a video search upgrade!

You’ll soon be able to search using videos. This means you can record a problem, like a malfunctioning record player, and Search will use AI to understand the video and provide solutions with steps and resources. No more struggling to describe the issue with text!

What does this mean for SEO?

The SEO Impact of Generative AI in Google Search

In my view, the question of how generative AI in Google searches will affect search engine optimization (SEO) has been a major concern. Is SEO dead? While the impact is undeniable, I believe SEO will adapt rather than become obsolete.

Firstly, website traffic might be affected. Users may get some answers directly from search results, reducing clicks to websites for basic information.

Secondly, content creators like bloggers and writers will face a challenge in crafting high-quality, relevant content that stands out for AI algorithms. However, Google’s inclusion of recommended articles within the search generative experience suggests that valuable content will still be crucial.

In essence, generative AI offers easier information access for users but presents a challenge for content creation.

What are your thoughts?