Astellas’s senior director, head of enterprise insights and digital solutions (EIDS), digital, analytics and technology, Masanori Ito weighed in on the promise of generative artificial intelligence (Gen AI), currently a “hot topic” not just for pharma but across sectors.

Ito believes that for now Gen AI can potentially augment the “creativity of employees” and assist them to find more generative novel solutions.

One of the use cases, he referred to is a kind of a “digital summarization”, given the plethora of documents required to be submitted to regulatory agencies for new drugs.(Also see "ChatGPT Unleashed: Generative AI Use Cases Taking Off In Pharma" - Scrip, 6 June, 2023.)

“So, there are so many opportunities to summarize and draft. Still, we need some hard manual process, but we can decrease the workload for them [employees],” he said.

In areas like pharmacovigilance, the technology can be used for text summarization or contribute to product safety information, he noted.

The company is also using large language models to contribute to the strategic decision making in a number of areas such as business development or portfolio strategy.

McKinsey in a recent report, said that it expects generative AI to have a significant impact on the pharmaceutical and medical-product industries - from $60bn to $110bn annually. Gen AI, the global management consultancy estimates could deliver potential value of 15-25% of operating profits for pharma when deployed across areas like R&D, content for commercial reps, etc.

Ito, though, also acknowledged the risks around patient privacy and data accuracy when using generative AI in pharma.(Also see "BioAsia 2023: Leaders from Novartis, Apple Talk Innovation, Tech, Data Privacy" - Scrip, 28 February, 2023.)

“One idea is to create an original large language model in the internal company and train the model based on 100% correct information/data. That’s one idea but we still need to do many things for that,” he added.

On whether Astellas expects to do that solely in-house or prefers alliances, Ito said: “Of course we are thinking about some partnering, but always we need to consider some balance between how we increase our internal skill and capability as well. So we are discussing now”.

Earlier this year, Mitsui & Co., Ltd. and chip giant NVIDIA joined forces for Tokyo-1 — a project that aims to turbocharge the nation’s pharma technology push including providing high-resolution molecular dynamics simulations and gen AI models for drug discovery.

The initiative features an NVIDIA DGX AI supercomputer that can be accessed by Japan’s pharma companies and startups, reports said at the time. Astellas was among the companies interested and hoped to accelerate its research with molecular simulations and LLMs for gen AI through NVIDIA BioNeMo software, one report said at the time, though Scrip could not immediately verify where things currently stand.

Astellas Exec On Tech Tools For Better Go-No-Go Decisions, Gen AI

More from Business

More from Scrip