Marketing and AI: A Case for Human Oversight

AI tools like ChatGPT and Claude have had a significant impact on the work being done in Marketing, particularly for digital agencies. It has also made it easier for anyone to claim to be a digital marketer. In an industry that is not regulated, this has resulted in a large number of “digital marketing specialists” who offer services at very attractive fees but use generic tactics that are fully reliant on AI with minimal to no human oversight. Unfortunately, without the human experience or expertise, this can result in wasted marketing investment by companies in strategies that are too generalized or lack the human touch that can make marketing so effective.

How AI can be used in Digital Marketing

AI tools are used across a variety of digital marketing tactics at agencies. There are the more simple uses like helping to organize thoughts into emails, getting help when looking for specific information, or organizing tasks and projects. They are integrated into the tools we use for our work and into the algorithms running the ads through platforms like Google Ads and Meta Ads.

For Search Engine Optimization (SEO), they are typically used to help create content and speed up the creation of SEO specific tags. They can also help to audit websites, do keyword research, and come up with SEO strategies for sites.

In Digital Ads, they can be used to create ad copy and are also integrated into ad campaign setup on major platforms. In some campaign types, the platform asks for a few basic inputs, and the AI handles the rest of the setup.

As we wrote in our Predictions for Marketing in 2026 article, they can also be used to create imagery for companies to be used within display or social media ads or on websites. Some websites have decided to fully leverage AI imagery throughout their site, as it is easy to generate the exact image setting that the business wants to demonstrate. Many digital ad platforms now also automatically generate images that can be used in ads in the platforms without being asked. All in the name of simplifying the ad process and the claim that it will help improve performance. Google has introduced its AI generator “Banana Nano” into Google Ads to help advertisers leverage a variety of campaign types they may not have been able to in the past.

We’re also seeing a rise in the number of “vibe coded” websites for small businesses, where the websites have almost completely been created by AI. Sometimes this is through platforms like Wix, where they have integrated their own AI tool. Other times it can be almost completely custom code from an AI tool.

Where does it go wrong?

An AI-led approach to digital marketing can result in a wide variety of issues. The biggest and most common one is content or ads that are too generic to compete in search results or against competitors. When so many marketers or companies are using AI to generate the same kind of articles or blogs, then there is nothing to differentiate the websites. This results in sites pumping out huge amounts of content for little to no return or traffic. From an ads perspective, generic ad copy is less likely to be clicked on, which is further compounded by signalling to the ad algorithm that the ad is not relevant to the target audience, which results in it being shown less or for clicks to cost more.

Strategies for SEO led by AI tools can also be generic or focus on areas that are of low impact since they do not have the full picture of the website performance and are only creating the strategy through their limited lens.

For Digital Ads, the biggest issues come from the fully automated campaign setups. These tend to include the platform “best practices,” which are usually best for the platform to make money, not best for the advertiser. For Google Ads, this may include the Google Display Network and Google Search Partners in a typical search campaign. Neither of these networks perform well for Search ads and result in high numbers of clicks and ad spend with little return of leads or revenue. There is also the introduction of “AI Max” into ad campaigns now. When creating automated campaigns with the AI tools, this is usually selected as a default. However, this also can result in very poor returns for the advertiser. We have seen it match the ads to irrelevant searches, create new ad headlines that are irrelevant or incorrect from old content it has found on the client website or infer from that content. There is also a setting that allows it to change the specified landing page if it thinks another one would perform better. However, landing pages are usually selected for a specific reason when the creation is led by a human.

AI imagery has a variety of issues. The main one is perception. AI images lack authenticity; the artificial feeling from them can have a strong negative perception by audiences. Websites and companies that rely on them too heavily result in a brand that does not resonate with potential customers. As mentioned in the Predictions for Marketing in 2026 article, Coca-Cola’s heavy use of AI in its Christmas ad was noted with disapproval by many of the people I know. While AI image generators have improved from the days of being unable to effectively generate fingers on human images, they are still susceptible to mistakes. Mistakes that can result in a negative perception of the brand.

“Vibe coded” custom websites can have many issues. One of the most significant can be security risks since AI code has been found to be quite buggy and can result in security vulnerabilities when used by people lacking development experience. It can also result in sites that look good on the one device specified by the person using the AI tool to develop them but have poor performance on other devices, such as mobile.

The Human Approach Still Matters

While AI has introduced many ways to speed up analysis and digital marketing work, it cannot replace the human element needed to really make digital marketing work for businesses. It is important to remember, especially with digital marketing tactics where the development of the work is very impersonal, that we are ultimately advertising to humans, not robots. And as such, we have the emotion and empathy required to create ads and content that will have an impact. Some of the ad and branding that has resonated with me the most has come from humans, not AI. Tushy’s use of humour in its ad campaign and product positioning has always amused me and anyone who I have talked to about it. All of that was human-led before AI chatbots were ever introduced.

In a sea of content generated by AI tools, it is the content that has a human perspective that will stand out. Content and imagery that provides something valuable to the customer or potential customer. The most memorable companies are those that we develop a connection with, companies that evoke an emotion. There is also so much good first-hand experience in companies that does not get leveraged when the content is generated purely by AI.

This isn’t to say that marketing and marketers should avoid AI altogether. But the most successful digital marketing comes from a human-led approach that leverages AI as an assistant, but always with human oversight.