TECHNOLOGY by Deborah Garry and Esther Poulsen Why AI needs a cultural check AI has maintained a high profile since it went mainstream in 2023. The phrase “artificial intelligence” was first coined in 1950, but public awareness and interest increased dramatically with the release of commercial generative AI products. AI-generated content has become a dominant force in storytelling, advertising, and political messaging, but its rapid adoption has outpaced accountability. Many of us in business love AI. We use it to create content, streamline operational and HR tasks, and save time. But what are the less tangible risks, besides the compelling concerns around cybersecurity, training and reskilling our workforce, use of AI platforms for hacking, invasion of privacy, and protection of corporate intelligence? The cultural impact of AI Bias & lack of reliability: AI is trained on heavily biased views sourced from 19th and 20th century material with a monolithic gender, cultural, religious and economic perspective. It absorbs, multiplies and replicates what it learns, then presents this as “learned truths.” This is a crucial issue for us as women and any under-represented community, and anyone concerned about the cultural impact of content that can be inaccurate, biased, misleading, and grounded in misinformation and stereotypes. Cultural understanding is highly complex, involving layers of subcultures that AI models often fail to portray or understand. Recent examples: ■ AI-generated videos that falsified the involvement of teens in threatening ICE agents resulted in death threats against the youth. ■ In healthcare, recent studies show that AI tools still routinely misjudge marginalized groups, downgrading women’s care needs and offering unequal treatment plans based on race and gender. ■ As a test before attending a Filipina Business Conference, where we both were to present on AI, we ran separate queries about Filipina women. The information that came back heavily portrayed them as nurses, nannies, and maids. Fake news and the accountability gap New research coordinated by the European Broadcasting Union (EBU) and led by the BBC found that AI assistants routinely misrepresent news content no matter which language, territory, or AI platform is tested. In June-July 2025, professional journalists evaluated more than 3,000 responses from ChatGPT, Copilot, Gemini, and Perplexity against key criteria, including accuracy, sourcing, distinguishing opinion from fact, and providing context. Key findings: 45% of all AI answers had at least one significant issue; 31% showed serious sourcing problems; and 20% contained major accuracy issues including hallucinated details. As more and more people source their news directly from AI assistants, the quality and accuracy of the information they find matters. The rise of inauthenticity and “AI Slop” Much of online content is now AI-generated and feels (and is) fake. This inauthenticity extends to product and service marketing, making it hard for consumers to trust what is real. The ease of AI leads to a lack of verification, resulting in “AI slop.” A lawyer used AI to find case samples and was thrown out of court because the cases didn’t exist. A crime-alert app spread false alerts when it misinterpreted ambiguous communications as confirmed criminal events. This negatively impacts brand trust and authenticity when AI drives messaging inauthentically or harmfully. There are few consequences for any organization (private or public) that fails to disclose AI content or misuses it. It’s not just a matter of inauthentic messaging; democratic discourse is threatened by disinformation campaigns and the erosion of trust in visual evidence. A major concern is the lack of “gates” to prevent deepfakes of living people, which has ramifications for everyone from job seekers to those navigating domestic disputes to political candidates. Elon Musk’s xAI and its chatbot Grok have faced extreme scrutiny, legal action and implemented policy changes in some–not all–countries due to explicit “deepfakes” of real people, including children. The problem remains pervasive and unresolved. Skills atrophy: Businesses may opt for “good enough” AI-generated content rather than paying for talent and expertise, resulting in generic, mediocre content and ultimately, a decrease in brand clarity, originality, and marketing ROI. Opportunity and proper use Marketing to and for multicultural customers generates 70% higher engagement rates than general market campaigns, and multicultural consumers are growing in population and purchasing power. Getting this wrong through cultural missteps can result in immediate and lasting brand damage. The “hallucinations” and biases in generative AI outputs result from the nature of their training data, their reliance on pattern-based content generation, and the inherent limitations of AI technology. Acknowledging and addressing these challenges is essential as AI becomes more integrated into decision-making. When crafting your message and using AI as an enhancer, consider these principles: ■ Authenticity as an opportunity: In a “world of mediocrity” created by AI, brands can stand out by being truly authentic. ■ Human oversight first: AI should augment, not replace, expertise on messaging through a lens of culture, gender and generation. Any AI-generated content touching on representation should be reviewed by people in those communities. 94 enterprising Women
RkJQdWJsaXNoZXIy MzA5OTEy