When it comes to diversity, equity and inclusion (DEI) training, the tone of conversations, the way the facilitator engages and the emotions experienced in the learning process are quintessentially human. But, with the speed at which artificial intelligence (AI) is advancing, will they always be? Will AI help or hinder organization’s DEI training efforts?
AI, by its very nature, has the potential to be a powerful ally in the context of DEI. Organizations are already using technology to enhance accessibility for employees with disabilities, eliminate biased language in job listings and address inequality in succession planning through identifying patterns in corporate data. There is no denying the added value that DEI-focused AI brings to the business world — and to training, with personalized learning paths, content recommendations and adaptive assessments.
But for personal, emotive and sometimes controversial topics like DEI, how is AI influencing our business realities? Like many things, there are both benefits and drawbacks. Let’s examine the pros and cons of this emerging technology in terms of DEI.
Potential Benefits of Leveraging AI
Realistic interaction.
Realistic interaction: Introducing AI enhancements to existing technology such as virtual reality (VR) or augmented reality (AR) means instructional designers can generate immersive and interactive DEI training experiences that look and feel like the real world (through AI-generated backgrounds, characters and audio). This helps to create opportunities for learners to actively engage in realistic scenarios, simulate difficult conversations, and practice inclusive behaviors in a safe and educational setting — where they can engage in learning without the fear of making mistakes or causing harm.
Inexpensive, speedy and global.
Video generation platforms that use AI technology enable training teams to create video content quickly and at a lower cost than traditional methods. Rather than hiring studios and actors for your video creation, tools like Vyond or Synthesia offer AI-generated people and audio that represent a diverse range of accents and languages, at a fraction of the cost. The swappable selection of visuals and audio allows for content localization and multilingual adaptation at the click of a mouse — an essential consideration for companies with international audiences.
Bias identification and mitigation.
We’ve all sat in a diversity training session playing stock footage bingo — the slide with the person in a wheelchair, the various hands in a circle the diverse team high fiving each other. But rather than learning design being this overtly manufactured experience, AI tools such as Perspective API and Clarifai can help identify and mitigate against biases in training materials and content. By analyzing text, images and videos, AI algorithms can flag potential biases in language, representation or in example case studies. This digital proofing resource enables organizations to review and modify content to ensure it is inclusive, respectful and representative of diverse perspectives.
Enhanced accessibility.
In the modern era of AI, DEI training teams have an incredible opportunity to not only talk about inclusive practices, but also to demonstrate it. Workplace training, and the way it’s offered, or the process by which employees are expected to learn, are not always the most congenial or inclusive. AI technologies can transform learner accessibility for those who seek it. For example, AI accessibility tools scan materials for screen reader compatibility and color schemes for those with visual impairments. Speech recognition tools convert spoken words into text and vice versa for neurodivergent learners who may have challenges with visual or verbal communication. This enhances understanding and enables individuals to express their thoughts and ideas more effectively than ever before.
Considerations Around AI Dependency
A lack of human connection.
Impactful DEI training is built upon complex social and cultural dynamics. In classroom-based training, DEI trainers read and respond to the reactions and emotions in the room. They adjust discussions based on increased interest or needed depth. They contextualize the local nuances and engage with learners’ diverse experiences through empathy, curiosity and open dialogue. It reaches beyond the human connection and moves into the sensitivity of behavior, context and organizational maturity in a way that AI systems may struggle with.
Programming bias.
AI algorithms are only as unbiased as the data they are trained on. Right now, tools such as ChatGPT and Google Bard receive data from a mix of various online resources including public and private organizations, government reports, research publications, surveys and studies. This means that, although this encompasses a wide range of diverse viewpoints, it can also include biased, discriminatory or offensive content. Let’s say you used AI to write your gender equality eLearning. If your AI tool pulled its data from biased or sexist language patterns, it would not be long before your content reinforced the stereotypes that women are better suited to nurturing roles or administrative tasks, while men are portrayed as more suited to leadership positions or technical roles: Not quite the training message you wanted. If programming biases are not addressed during design, and your AI-generated content is not verified by a subject matter expert, at best you disenfranchise your corporate values. At worst, you implode your future talent pipeline.
Misinterpreted translation.
AI translation systems have come a long way, but they still have limitations in accurately capturing language nuances, cultural context and sensitivities. There are many words in the DEI vocabulary that do not exist in certain languages — for example, inclusion does not translate directly into Japanese. There are also words that do exist, but have vastly different meanings – for example, there is a certain Dutch term that causes particular offence for German speakers. While AI translations can provide a general understanding of the content, they may not always capture the intended meaning or convey the messaging in the most culturally appropriate manner. It is critical to be cautious and verify automated translation accuracy, especially for sensitive topics related to diversity and inclusion.
In Summary
Advancements in AI technology continue to improve the ability to mimic human behavior. However, when it comes to DEI evolution, the role of human specialists and their ability within such an emotive topic remains crucial. By combining human expertise and AI technological support, organizations can transform their DEI learning strategy to reach further — and with greater impact — than ever before.
