The recent crackdown by the Securities and Exchange Commission (SEC) on false claims of using artificial intelligence (AI) by investment advisors and broker dealers highlights the growing issue of AI washing, where companies exaggerate their use of AI to deceive investors and customers. This phenomenon is not limited to the financial sector, as analytics firm FactSet found that 179 S&P 500 companies mentioned AI in their earnings reports in the past three months, more than double the average of 73 in the previous five years. However, it’s unclear which of these companies are truthfully implementing AI technology.
Experts warn that AI washing leads to a loss of trust between vendors, consumers, enterprise partners, and investors. Companies may claim AI capabilities without actually having them, or may rely on simplistic applications that provide short-term benefits but fail to deliver long-term results. For instance, when a law firm replaces a human with an AI assistant, it may stop working effectively if the AI lacks a unique and unbiased dataset specific to the legal field and fails to continuously learn.
To avoid falling prey to AI washing, industry watchers recommend that businesses clearly define what type of AI they use and explain how they implement it. They should also indicate what they won’t do with AI and adopt an ethics policy. Moreover, investors should scrutinize startups using the four D’s criteria: data availability, dividend contributions, distribution channels, and delightful user experience. As AI washing becomes increasingly common, regulatory bodies are expected to intensify oversight and impose fines against offenders.
Microsoft’s venture capital arm M12 employs these principles in evaluating AI startups. The firm assesses whether the startup has exclusive access to critical data required for AI operations and whether AI contributes significantly to the company’s revenues. Additionally, M12 considers the startup’s distribution channels and user experience. In the past, M12 eliminated an entire category of computer vision firms due to insufficient evidence of AI implementation.
Overall, the AI landscape is fraught with companies exaggerating their AI capabilities, resulting in a loss of trust among stakeholders. To mitigate AI washing, industry leaders recommend that businesses clarify their AI applications and adhere to appropriate standards while investors need to be cautious and conduct thorough investigations before investing in AI-based ventures.
Leave a Reply