RedPeg Logo
a
a
a

Avoiding the AI-Pocalypse: Using Artificial Intelligence Ethically

June 2023
  • The adoption of AI by brands across different industries is increasing at a rapid rate
  • Despite still being in the infancy of AI development, brands are using AI because it is positively impacting business efficiency, especially in marketing, where machine learning can transform consumer data into personalized marketing
  • AI in its current form has some ethical concerns, particularly when it comes to data privacy, built-in biases, and copyright claims, all of which must be carefully considered by brands before integrating AI
  • As one of the last remaining trusted institutions by the general public, brands have a heavy responsibility to carefully and safely navigate this new technological landscape

 

INTRODUCTION

The advancement of artificial intelligence (AI), and its usage has grown exponentially around the globe. Since 2017, AI adoption in businesses has increased 250%, and now half of all businesses use AI in at least some capacity. You’ve probably seen and heard AI as a massive talking point over the last couple of years, and rightfully so. Not only is it part of the cultural zeitgeist, but it is also a critical question for businesses and brands as to how and where they should (or shouldn’t) be using AI. In this Direct Hit, we take a deep dive into how to protect your brand by using AI strategically and ethically, as AI capabilities and its availability continue to increase.

 

Since 2017, AI adoption in businesses has increased 250%, and now half of all businesses use AI in at least some capacity (courtesy of McKinsey).

 


THE GOOD: WHY BRANDS ARE BUYING INTO AI

While the movies can make AI seem like a scary technology to avoid (looking at you, Skynet), there are many reasons why brands are increasingly integrating AI into their capabilities. The three biggest reasons brands are using AI: 

  1. Efficiency: Perhaps the single biggest reason that brands are buying into AI is the potential for massive growth in business efficiency. Instead of having to do a task manually, or supervise and employee doing a task manually, once you set up an AI tool to do the task, it will do it autonomously. This frees up time that your organization would have otherwise spent on the task, and is particularly useful for menial, repetitive tasks. As AI capabilities increase over time, there is also the opportunity to do these kinds of tasks at scale, exponentially boosting efficiency by being able to automate many tasks simultaneously.

     

  2. Personalization: Since AI is based on machine learning algorithms, which at its simplest involves processing and ‘learning’ from large amounts of data and applying that learning to new data, it allows for brands to personalize individual content in real time and at scale that was previously impossible. This is particularly useful for digital content and marketing outreach, where AI can tailor images and messaging to individual consumers based on their preference data. Not only is this more efficient, but far more optimal from a customer conversion and acquisition standpoint.

     

  3. Revenue Gains: With increased efficiency comes revenue growth. Across different industries, the adoption of AI has led to, on average, 63% of businesses seeing at least 5% revenue increases. Marketing is capitalizing on this, where that number jumps up to 70% of businesses. On the flip side, we’re seeing an average cost decrease of at least 10% in 32% of businesses that have adopted AI. The adoption of AI is directly facilitating greater revenue gains while simultaneously cutting down on business costs.

     

Graph depicting cost decrease and revenue increase from AI adoption in 2021 (courtesy of McKinsey).


THE (POTENTIALLY) BAD: POSSIBLE ETHICAL CONCERNS SURROUNDING AI

Sometimes the movies, while being theatrical exaggerations, can point out some of the potential warning signs to pay attention to with new technologies. In the case of AI, which is relatively new in wide-scale adoption, there are some elements that brands should be mindful of before blindly adopting it.  

  1. Data safety and privacy: Particularly in marketing, content personalization requires large amounts of consumer data to collect and process. With this comes the risk of being perceived as invading consumer privacy, which can not only lead to legal concerns, but also decrease consumer trust in your brand. If you’ve been following the Direct Hit, you’d know just how important brand authenticity & transparency is for consumers today. It’s even more important when you realize that business is the only trusted institution left for consumers, so risking that trust is an existential threat for brands.

     

  2. Model bias: AI models need data to train on. What that data looks like is very important; if the input data includes bias, so too will the output. This is especially a concern for any AI that trawls the internet for data, as bias and misinformation continue to run rampant on the internet. Take Amazon’s new AI facial recognition software, ‘Rekognition’, as a potential warning. A recent test of the software disproportionately falsely matched POC members of Congress with criminal mugshots, because the model was trained with mostly white faces, leading it to be less accurate in classifying POC faces. Even OpenAI, the company behind the wildly popular ChatGPT, openly states in their best practices document that “From hallucinating inaccurate information, to offensive outputs, to bias, and much more, language models may not be suitable for every use case without significant modifications”. Why does this matter for brands? Because without human oversight and verification, there is a danger that fully automated AI marketing tools will output offensive, biased, or incorrect information directly to consumers, significantly damaging your brand’s image and consumers’ trust.

     

  3. Copyright infringement and non-credited content use: AI content generators trawl the internet for millions of pieces on input data and can end up generating content that is nearly identical to specific artists’ styles. This is currently a legal gray area, but there is a significant possibility of legislation against this that could end up leading to major lawsuits. In fact, there is currently a pending class action lawsuit against GitHub Copilot for this exact reason. These possibilities have two major implications for brands here, with the obvious one being that one could accidentally open their brand to legal action for copyright infringement. The other, however, is alienating your own creative teams, who likely have their own ethical viewpoints and concerns over unvetted AI content generation. No creative wants their work to be copied, directly or indirectly; and how your brand addresses those concerns could be critical for keeping your organization aligned internally. 

People of color were disproportionately falsely matched in a test conducted by the ACLU (courtesy of the ACLU).


CONCLUSION

As we’ve explored, there’s a lot to be excited about when it comes to AI, but there are also many ethical implications surrounding when and where it should be used, especially when it comes to marketing initiatives that are public facing. While AI in marketing can significantly increase efficiency and revenue, there is also the risk of replacing human intuition and creativity that is fundamental to impactful marketing, resulting in potential legal trouble or negative consumer reactions. As AI continues to develop and become more powerful, brands have the responsibility to ensure that they harness this technology ethically and consider its implications on brand image, safety, and consumer relationships. Apocalypse avoided, for now. 

Want more of The Direct Hit?