On The Ground At BPI: Kat Kozyrytska On Ethical AI And Collaboration

It’s almost impossible to escape the topic of AI at any biotech conference these days, and I’m sure that won’t change anytime soon. For better or worse (you decide), AI is here to stay, and that presents as many challenges as it does opportunity for the industry. Perhaps the biggest challenge — next to finding effective uses — is figuring out how to ensure compliance and ethical AI use in this highly regulated industry.

The latter point was the focus of a session at this year’s BioProcess International (BPI) show in Boston titled “Leveraging Irresistible Efficiencies: AI Deployment Framework To Protect IP And Ensure Compliance And Ethics,” presented by independent AI consultant Kat Kozyrytska. While I couldn’t attend her talk, I was able to connect with Kozyrytska and ask some questions for our audience.

The following transcript has been edited for clarity.

How has your time at BPI been so far? What are your goals for attending the conference?

Kozyrytska: I’m here to embrace all of the learnings from the work various companies have done in the past year. I think there's so much on the digitization and application of AI. We had an amazing keynote from Thomas Seewoester, Ph.D., at Amgen, where he talked about this — for some companies, futuristic — concept of hybrid organizations where you have humans collaborating with AI agents. For a lot of other sectors, that’s a reality, so I think we have to embrace the idea that these are here, and it's a today situation. It's not really something we should attempt 30 years from now.

Your talk is on AI deployment and IP protection. What do you hope audience members take away from your session?

Kozyrytska: In AI, I think there are so many opportunities that we have, but there are also so many risks. I see a very interesting and different dynamic in the AI space compared to some of the other technologies that we've implemented. I work across the ecosystem with therapy developers, AI technology providers, and regulators to help us get to a good level-set on the certifications and standards that have to be in place for this technology to become useful, while maintaining ethics and morality within our space, adherent to our professional code of conduct.

I think on the opportunity side, it's getting the insights within your own dataset, while also unlocking the opportunity to collaborate in privacy- and confidentiality-preserving ways that we were never able to do before. That collaboration, as exemplified by Eli Lilly and the launch of TuneLab last week, opens up an opportunity for us to start innovating at a much lower cost and on a much faster timeline. Hopefully, that will get investors really excited in our sector again, because compared to other sectors, we can really accelerate by working together and analyzing these data sets with AI-powered analyses.

Do you think we’re any closer to this collaboration you’re describing?

Kozyrytska: I feel very strongly about this. Decentralized, AI-powered analyses platforms have been around for a while, and the decentralized science movement has been around for a while, but I think the work by Eli Lilly and that collaborative project sets a new stage and gives us hope that we can replicate this in other sectors.

I think the biggest opportunity for financial impact potentially lies with novel modalities. We can still think of cell and gene therapies as the newer space, but understanding the drivers and predictors of cell therapy manufacturability through these kinds of AI-powered collaborations I think will help us move so much faster.

What are your biggest takeaways from this Biotech Week? What sessions or conversations were you excited by?

Kozyrytska: I am completely biased, but I think Tracy Ryan's keynote was so patient-centric, about her experience as the mother of a young child with cancer and how they handle that as a family. I mean, it's just so profoundly impactful, and it helps us bring back that patient centricity and really focus on patient outcomes. Of course, we have to be operational, right? We are businesses. We need to keep the ROI in place, but I think really focusing on the impact for the patients can help us work together to get better outcomes for these patients, faster.

What do you want to let the Bioprocess Online audience know about AI or technology adoption?

Kozyrytska: Just an add-on to what we were talking about before with these hybrid organizations. Within our regulated sector of pharma, biotech, life sciences, and healthcare, we expect the humans that we employ to adhere to a very strict ethical code, part of which is confidentiality, and another part of which is acting in the patient's best interest. So, for the AI that we're building within this sector, the requirements on that front are so much higher than for dinner-suggestion AI, because there is a certain level of expectation of ethics, and I would even argue, morality. Not just the external rules, but the internal compass for what's the right thing to do. And so, I think implementing that thinking as you are bringing AI into your organization, to ensure it's serving your organization's best interests and increasing productivity, is going to be key moving forward.