Understanding biopharma's AI integration strategy: What's the plan?
SUMMARY
Biopharma leaders and researchers are focusing AI integration on training, pilot projects, and partnerships
Over ⅔ of our cohort acknowledge challenges in establishing a robust framework for the practical adoption of AI technologies
Half of respondents anticipate AI implementation to result in job increases
Leaders are focusing on certain key performance indicators to demonstrate the utility of AI adoption
OUR AUDIENCE
We captured the sentiment of 25 scientists, directors, and executive leaders with varying levels of influence in purchasing and implementation of AI tools from biotechnology companies, biopharmaceutical research/development companies, contract research organizations (CROs), and pharmaceutical research/development companies spanning all org sizes from small (<100 employees) to enterprise (10k+ employees).
To meet productivity demands this year, biopharma is turning to AI and automation. But in practice, what does adoption of AI tools really look like for this audience? And where does biopharma see the steepest challenges to overcome and the most robust opportunities for success?
A cross-functional mindset
Biopharma is approaching AI implementation with a cross-functional mindset, often citing multiple measurable objectives beyond purely scientific application. Areas of opportunity include:
Data analysis or bioinformatics across the organization (mentioned by a majority — nearly ¾ of leaders)
Process development, scale-up, optimization (mentioned by over ⅔ of leaders)
Target identification (e.g., predicting molecules) (mentioned by over half of leaders)
As biopharma integrates AI, certain tasks like data-related tasks, administrative and regulatory tasks (where AI is seen to make the quickest impact in creating and managing documents), facilitating pharmacovigilance, optimizing supply chain management, and drug discovery and target identification will be the fastest areas to see implementation.
Diving deeper, biopharma’s specific needs look like:
Full Stack AI and Machine Learning Integration
Combining Data from High-Throughput and Novel Technologies
Customized Solutions bespoke to workflows and operations
Where leaders are focusing
Speaking to leaders, we heard a resounding need to focus on training, pilot projects, and partnerships. Related to training, leaders are focusing on internal knowledge deployment to build expertise and successfully integrate AI throughout the entire organization. Pilot projects give leaders the ability to test and refine how certain programs and projects may not only benefit from AI, but also understand where AI may not be the right fit for efficiency.
Over ¾ of leaders are seeking or plan to seek outside expertise (through service providers, consultants, technology providers, or partnerships) to help with AI-related implementation challenges. This highlights unique opportunities for vendors if they can demonstrate their commitment to working with external service providers and adopting at a rate that matches internal capabilities. Vendors also have an opportunity to ensure all AI applications are well-regulated and legal, particularly related to proprietary scientific literature and copyright issues.
Confidence in AI adoption prioritizations
Although 60% of respondents express contentment with the prioritization of AI in their organizations, a notable proportion (36%) acknowledge challenges in establishing a robust framework for the practical adoption of AI technologies. Key barriers cited include:
Cultural resistance to change
Lack of expertise and infrastructure
Data credibility and successful case studies
AI’s impact on employment
A critical area to consider is how AI and automation will impact internal staffing resources and employment. Many leaders are looking closely at how AI may increase or decrease staffing needs. Half of the researchers and leaders we spoke to anticipate AI implementation to result in job increases within three years. That anticipation is due to AI’s increased data output, necessitating new specialists in AI departments, proteomics, IT, R&D, and data science. That also includes anticipated needs in roles across biology, chemoinformatics, and genomics that support data management and AI training.
Only 16% of the leaders we spoke to expected a decrease in jobs due to AI, particularly in IT, QA, regulatory and patient safety functions. AI-induced process automated systems will also impact customer support and inventory management.
AI tools’ impact in laboratory experiments
Novel data and wet lab experiments are still seen as essential for target validation, construction and testing of models, and in-depth exploration of disease processes. However, there’s hope that AI may also enable more efficient methodologies through enhanced automation, increased accuracy, improved predictive analytics, and optimized upstream processes.
How will success be measured?
Driven by metrics, leaders are focusing on six key performance indicators to demonstrate the utility of AI technology adoption:
Operational Efficiency and Workflow Acceleration
Innovation Yield and Scientific Breakthroughs
Precision and Error Reduction
Ongoing Improvement and Adaptive Integration
Benchmarking Strategic Performance
Financial Impact and Cost-Effectiveness
With this field moving quickly and pharma feeling pressure to make significant progress in 2024, we’ll be keeping a close eye on consumer sentiment and partnership developments. And, in just a few short months, we’ll begin collecting data for our H2 2024 State of Science survey, which will include assessments of progress and updates on the insights from this study.