HPE CTO on Using AI with Guardrails: Focus on the Right Use Cases and Monitor Data Use

Fidelma Russo, Chief Technology Officer at Hewlett Packard Enterprise (HPE), has over 30 years of experience in technology leadership. She notes that while HPE has invested in AI for many years, recent advances have created an inflection point where previously intractable problems can now be addressed.

In an interview, Russo discussed HPE's approach to data, AI's potential impact across the enterprise, and the need for close governance. She explained that AI is already powering many HPE products and operations through machine learning. HPE's supercomputing division is focusing on advances like generative AI, large language models, and tools to help developers. AI is also being applied to supply chain, support, and software development to speed processes, increase predictability and productivity, and improve customer satisfaction.

While exciting, Russo emphasizes the need for maturity to manage this transition responsibly. For example, HPE is testing AI's effectiveness in customer service through ongoing pilots. Partnerships are exploring how models can be built and trained with internal and partner data to resolve issues faster.

Most importantly, HPE established a company-wide AI oversight board to guide responsible data use and model development. The board monitors for potential bias, ensures privacy protection, and aligns results with expected outcomes. It also considers sustainability innovations. With rapid data growth and need for continual model training, responsible and sustainable data practices are an ongoing challenge.

The oversight board includes executives from technology, legal, cybersecurity, data governance, and other functions. Its cross-functional composition allows consideration of AI's effects across the business. The board sets policies and oversees working groups exploring beneficial AI uses and processes.

Russo's advice is to view AI as relieving tedious tasks and delighting customers, rather than fearing problems. While generative AI breakthroughs may be limited by training on existing data, the technology is already transforming work and products. The time has come to enable AI responsibly through governance discussions. Success comes from the right use cases, not hype. Overall, AI promises exciting opportunities if approached with care, oversight and human judgment.