Validity and reliability Security Security and resilience Accountability and transparency Explanation and interpretation Privacy Fairness by reducing harmful bias To examine today’s responsible AI landscape across enterprises, MIT Technology Review Insights surveyed 250 business leaders on how to implement principles that ensure AI reliability. The poll found that responsible AI is important to executives, with 87% of respondents rating it as a high or medium priority for their organizations. The majority of respondents (76%) also said that responsible AI is a high or medium priority specifically for creating a competitive advantage. But relatively few have figured out how to turn this idea into reality. We found that only 15% of those surveyed felt very prepared to implement effective responsible AI practices, despite their importance. Implementing responsible AI in the age of generative AI requires a set of best practices that leading companies are adopting. These practices can include cataloging AI models and data and implementing governance controls. Companies can benefit from conducting rigorous assessments, tests, and audits for risk, security, and regulatory compliance. At the same time, they must also empower employees with training at scale and ultimately make responsible AI a leadership priority to ensure change efforts are sustained. “We all know AI is the most impactful change in technology we’ve seen, but there’s a big disconnect,” said Steven Hall, chief AI officer and EMEA president at ISG, a global technology research and IT advisory firm. “Everyone knows how transformative AI is and wants strong governance, but the operating model and funding provided for responsible AI is below par, so it’s worth criticizing these organizations.” Download the full report. This content is produced by Insights, the dedicated content arm of MIT Technology Review. It was not written by the MIT Technology Review editorial staff.
