Since the introduction of advanced large language models (LLMs) like ChatGPT and image generators like Stable Diffusion into the mainstream, the marketing industry has undergone a significant overhaul. AI-generated content has become increasingly popular with marketers, and we’ve seen the proliferation of AI avatars—artificially generated representations of brands and personas.
Over the past year, these AI avatars have become increasingly sophisticated. Their role in brand marketing has changed from novel sources of entertainment to true brand ambassadors. Major companies like TikTok have introduced AI avatars as customer-facing brand representatives and as customer support personnel. However, the transition has come with its challenges. Namely, brand authenticity is taking a hit, and the consumer reaction hasn’t been everything that companies had hoped for.
Authenticity
There has been a lot of talk recently in communications circles about how authenticity has become a rarer and more important commodity as more brands are churning out generic, impersonal AI-generated content. Similarly, if the majority of interactions your customers have with your brand are with digital representatives instead of real humans, customers may view your company as artificial and phony.
Also, if AI-generated content is not clearly labeled as such, when customers realize they’re interacting with an AI avatar rather than a human, they may feel deceived. This can lead to a perception that the brand is trying to cut corners or avoid genuine engagement with its audience. For example, in April some of Meta’s chatbots started posing as human on social media, leading to confusion and frustration from users.
It’s true that AI avatars lack the level of emotional intelligence and empathy that human representatives can bring to customer interactions. Some customers may feel disconnected or undervalued, particularly in the process of resolving complex issues or providing support in stressful situations. In one study, nearly half of respondents said they preferred to talk to a human over a chatbot for customer support, while just 12 percent of respondents strongly preferred using AI chatbots.
Lastly, companies need to contend with potential bias and other errors when using AI avatars as the face of their brands. Without proper training and monitoring, AI avatars can perpetuate or even exacerbate biases. This can occur in the form of inappropriate language, stereotyping in interactions, or unequal treatment of customers based on demographics, which could lead to public backlash and damage the brand’s reputation.
For example, recent studies have shown that many large language models perpetuate biases and stereotypes about speakers of African American English (AAE). AI avatars can also simply give incorrect or misleading information or fail to understand complex queries. These failures can be particularly damaging if they occur in high-stakes environments like financial services or health care.
Best practices
Of course, there are benefits to AI avatars that make addressing these challenges worth the effort. First, AI avatars can offer a more approachable brand presence and a more personalized and memorable experience for customers.
Let’s start with the approachability aspect. AI avatars don’t have our human tendency to get annoyed with repetitive questions. They’re designed to be unerringly friendly, engaging, and accessible at any time. The constant availability and rapid response time can make brands appear more approachable, especially to audiences that are used to instant gratification.
When it comes to personalization, AI avatars can remember previous user interactions, preferences, and even mood to make responses and recommendations more specific to each customer, so interactions are more relational and memorable. Companies can also tailor AI avatars’ appearance, language, and behavior to particular demographics, ethnicities, or cultures.
When you’re using AI avatars, it’s important to always maintain a consistent brand voice, since major inconsistencies in communication styles and content can confuse or alienate the audience. Fine-tune your avatars with a deep understanding of the brand’s personality to make sure that every interaction aligns with how the brand communicates across other platforms, whether the style is professional and tech-savvy or more casual.
Also, never try to pass off AI-generated content as human-made, as this can lead to negative perceptions and seriously erode trust. Consumers shouldn’t have to feel deceived about whether they are interacting with a human or an AI. Clearly label AI-generated interactions as such to allow consumers to set their expectations accordingly.
Be sure to include disclaimers and reporting mechanisms so that customers can flag errors, bias, or otherwise unusual content. The platform could feature a “Report an Issue” button that lets users flag any problematic interactions, with a promise of a prompt review and response. If the AI avatar has known constraints, such as difficulty understanding complex queries or providing nuanced advice, you should communicate this to users upfront.
For example, a financial services company might use an AI avatar to provide basic information about account management. However, the avatar could include a disclaimer stating that if customers are seeking more complex financial advice, they should do so with a human advisor.
AI and the future of brand representation
As AI technology continues to grow, the role of AI avatars in brand marketing will also expand significantly. AI avatars will become more integral to content creation and delivery, managing social media interactions, and even participating in real-time outreach and marketing campaigns.
However, those benefits must not come at the expense of brand authenticity. And with the right best practices in place, they won’t have to.
