Why AI Feels Like the Wild West
The Showdowns Defining AI in Life Sciences Today
The rapid expansion of the American West created a landscape defined by opportunity, competition, and uncertain rules. Today, many life sciences leaders may feel they are navigating a similar frontier with artificial intelligence.
The AI territories for industry transformation look both extensive and attractive: drug design, trial simulation, data cleaning, regulatory writing, quality monitoring, employee self-service, and many more paint a vibrant picture of AI-enabled businesses that deliver better therapies faster. But the day-to-day experience can sometimes feel like a showdown of competing concerns:
Velocity vs. Planning. Frontier AI companies are releasing product updates at an unprecedented pace – measured in months, not years. Combined with the continuous influx of new AI-enabled players in the market, corporate AI roadmaps become outdated almost as quickly as they are authored.
Use vs. Proliferation. Industry leaders want to encourage the advancements and efficiencies that AI can bring (e.g., faster recruitment, streamlined study operations, cleaner clinical data). But AI growth brings practical challenges: proliferating data copies, rising compute costs, tool sprawl, and increasing compliance risk.
Agility vs. Governance. From the board room to the break room, leaders and workers are embracing AI — from tools that summarize clinical trial documents to copilots that draft regulatory submissions. As organizations struggle to make decisions and implement new solutions, research shows a large portion of employee adoption is “shadow AI” use of ungoverned and unsecured products.
Market Enthusiasm vs. Value Creation. The market noise around AI is deafening, and expectations are frequently misaligned to reality. Leaders struggle to balance the imperative to move with a disciplined strategy that focuses on the most attractive and measurable value creating opportunities.
Duplication vs. Standardization. Practically every IT vendor in every organizational unit is bringing AI to market. Leaders responsible for quality deliverables, compliant operations, and controlled costs are grappling with a deluge of independent models offering similar capabilities, differing reliability, and growing costs.
Build vs. Buy Decisions. Provisioning AI capabilities is not a simple “build vs. buy” equation, and thanks to AI, tailored solutions don’t necessarily require manual coding. Though AI tools can be purchased “off the shelf,” more impactful AI solutions that deliver higher value for any organization are often some combination of both.
Change vs. Quality. Business transformation with AI requires workforce education, process re-engineering, and evolving roles. In the absence of disciplined programs focused on effective transitions, the corresponding AI capabilities frequently face problems in accuracy, scalability, compliance, trust, adoption, and ROI.
How the West is Won
The AI frontier may feel chaotic, but organizations that succeed are not improvising. They operate with a clear set of principles that guide technology, governance, and change.
One Program, not Many Projects. Scalable AI capabilities with measurable value creation do not emerge naturally from grass-roots efforts. As opposed to waiting to uncover small-scale wins, create an orchestration center that can drive a more managed approach across the business. It is faster and cheaper to create the right investments once.
Strategy, not Roadmaps. The AI landscape will not stabilize any time soon. Instead of trying to maintain static adoption roadmaps that are always out of sync with the market, pivot to focusing on clear strategies, technical principles, and change-ready blueprints that guide the ongoing selection and migration of AI capabilities over time as the business and technology evolve.
Data, not Tech. Most risks in AI projects have nothing to do with the AI itself. Experience shows that most organizations do not have data assets designed to perform well with AI (e.g., semantically consistent research data; managed master data; data architectures that effectively blend structured and unstructured data across scientific, operational, and financial domains). Instead of evaluating the next great AI tool, direct teams to carefully assess the quality, consistency, management, and trustworthiness of the organization’s key data that would be used by any AI model selected.
Quality System, not Documentation. The organization’s obligation to cultivating and using high-quality information systems – whether regulated or not – grows with AI. When systems are not deterministic, technology is changing, and processes are being re-engineered, it is critical to address quality as a comprehensive framework guiding effective, safe, and compliant business operations.
Enablement, not Deployment. To unlock sustainable value creation with AI, adoption is usually a transformational journey. The goal is not to give employees more software. Focus programs on enabling employees to perform different processes, empowering them to co-create the future, leveraging new skills, and freeing capacity from manual labor that machines can undertake.
The AI frontier in life sciences will not normalize anytime soon. But history suggests that periods of technological expansion eventually reward the organizations that combine ambition with discipline. In the AI era, the winners will not simply adopt new tools faster — they will build the strategies, data foundations, and operating models that allow those tools to create lasting value.
