In early 2023, Microsoft found itself at the centre of an AI controversy after launching an AI-powered chatbot within its Bing search engine. The move, part of the company’s broader strategy following a multi-billion-dollar investment in OpenAI, was meant to showcase how artificial intelligence could transform everyday technology. But when the chatbot started behaving unpredictably—offering unsettling conversations and even using offensive language—the tech giant faced widespread criticism.
Microsoft acted swiftly, limiting the chatbot’s responses and eventually replacing it with “Copilot,” now integrated into its Microsoft 365 suite and Windows operating system. The incident, however, raised questions about the risks of deploying AI without addressing underlying biases—a challenge Microsoft and the wider tech industry continue to grapple with.
Critics have long warned that AI systems can perpetuate racial, gender, and other biases because they are trained on massive datasets pulled from the internet. These data often reflect the inequalities already present in society. One infamous example involved Google’s Bard tool, which made a factual error during a live demo, causing a significant loss in the company’s market value. Later, it faced backlash for allegedly refusing to generate images of white people in response to specific prompts.
A Push for Diversity in AI Development
Microsoft believes that increased diversity within the teams designing AI can help mitigate these biases. Lindsay-Rae McIntyre, the company’s Chief Diversity Officer, emphasises that creating “inclusive AI” is now a top priority for the tech giant. McIntyre, who joined Microsoft in 2018 after two decades in human resources at IBM, insists that diversity is key to ensuring AI’s future fairness and inclusivity.
“We’re embedding inclusion practices into all levels of AI research and development,” McIntyre says, noting that diverse voices within development teams lead to more empathic, accurate, and relevant AI technologies.
Microsoft’s commitment to diversity is part of a broader shift within the company, which reported that nearly 55% of its workforce comes from racial and ethnic minorities. While progress has been made, there’s still room for growth—particularly in gender diversity, where women make up just over 31% of the workforce, trailing competitors like Google and Apple.
AI Bias: A Systemic Problem
Despite Microsoft’s optimism, experts argue that simply diversifying teams won’t entirely eliminate AI bias. Large language models like those used in ChatGPT, Copilot, and other AI tools rely on training data sourced from the internet, which can be rife with prejudice. This makes it difficult to filter out bias completely, especially as AI systems become more embedded in real-world applications.
For instance, recent studies have shown that AI systems often associate certain professions with specific genders, reinforcing stereotypes. Jobs such as “flight attendant” and “secretary” are frequently categorized as feminine, while roles like “judge” and “lawyer” are labelled as masculine. This presents challenges for industries that are increasingly relying on AI, including in critical areas such as hiring and law enforcement.
Still, Microsoft remains optimistic that AI can be part of the solution if built with inclusion in mind from the outset. The company’s recent innovations, such as the Xbox Adaptive Controller for users with disabilities and accessibility features across Microsoft 365, highlight its long-standing commitment to inclusivity.
Localising Diversity Initiatives Globally
Microsoft’s focus on diversity goes beyond its US headquarters. The company has rolled out initiatives across different regions to address specific cultural needs. In India, for example, Microsoft has tailored its diversity training to include discussions about the intersections of race, ethnicity, and religion. Meanwhile, in Australia and New Zealand, the firm is working with Indigenous communities, while in the Middle East and Africa, Microsoft has launched support programs for women experiencing menopause, an issue often neglected in workplace policies.
According to McIntyre, the localisation of diversity strategies is crucial to ensuring that employees in all 190 countries where Microsoft operates feel represented and supported.
AI in Human Resources: A New Frontier
One area where AI is making a significant impact is within Microsoft’s human resources department. Copilot is already being used to answer employee queries more efficiently, while AI-powered skilling courses are helping workers across the company stay up-to-date with the latest technology. McIntyre believes that other organisations can benefit from a similar approach, noting that “AI can help HR teams foster inclusion and empathy” if used thoughtfully.
For Microsoft, the future of AI is both a challenge and an opportunity. By embedding diversity into the core of its AI research, the company hopes to avoid the missteps of the past while advancing towards its goal of creating technology “that understands us,” as CEO Satya Nadella has put it. Whether this vision can be fully realised will depend on the company’s ability to continuously innovate while keeping inclusivity at the forefront of its AI journey.
