The Pistoia Alliance’s recent findings highlight escalating concerns regarding artificial intelligence (AI) in life sciences, especially related to data provenance, licensing, and the visibility of scientific content in AI models. Over 25% of life sciences professionals are unaware of the data utilized by their AI systems, often relying on abstracts rather than full-text resources. This discrepancy underscores a growing divide between complex research needs and the reliability of current AI models. A survey of over 170 professionals revealed that only one-third integrate internal documents into their AI systems, resulting in reliance on incomplete scientific evidence. Issues of unclear governance and inadequate licensing practices expose organizations to compliance risks. To combat these challenges, stronger standards for benchmarking AI systems are essential. Experts agree that successful AI adoption hinges on combining technology with skilled personnel, emphasizing the necessity for collaborative efforts to cultivate trust, transparency, and robust data quality in AI-driven research.
Source link
Share
Read more