The transformative potential of artificial intelligence (AI) is no longer just a topic for future planning—it’s an imperative for today’s life sciences leaders. And yet, despite the tremendous amount of industry attention and hype being directed at AI, the path to value creation with AI is not clear or paved.
Investors, board members, leaders, and employees hear about the potential of AI daily. Company executives are increasingly feeling pressure to describe their strategies and plans, but many are unsure how to take charge of AI adoption in a way that not only aligns with their organization’s goals but also advances meaningful business and clinical value.
AI roadmaps do not need to be big or risky. I’ve compiled a list of eight popular and well-considered ways that industry leaders can start making progress in the application of AI across their organizations. Some are simple, even optimizing capabilities you might already have. Others are designed to facilitate AI-powered breakthroughs, catalyzing new thinking about how your organization can embrace the transformative potential of AI. But each one offers an opportunity to deepen your company’s perspectives on how to grow with this exciting technology.
1. Establish AI Guardrails
For many life sciences leaders, one of the first priorities with artificial intelligence is defensive – protecting the company’s information. At the outset of your company’s AI journey, developing robust policies and comprehensive employee guidance around the use of AI is helpful in safeguarding intellectual property and confidential information. Many employees are not familiar with the risks of AI technologies – exposure of confidential information, inaccurate search results, potential exposure to information rights claims by third parties, evolving regulations, and more – that should inform how they leverage AI systems.
Guardrails such as AI policies actually serve two purposes: protecting the company’s interests and educating employees about AI issues that need to be top of mind. The goal should not be suppressing the use of AI. Rather, the goal should be empowering the safe, effective use of AI technology while minimizing risks. By proactively addressing these aspects, companies can foster a secure and compliant environment that supports innovation while protecting their people and valuable assets.
2. Request an Executive Briefing
Artificial intelligence is a broad and diverse field spanning computer hardware, data, computational modeling, software solutions, training, and many other disciplines. From the largest technology players in the world to start-up companies that did not exist a year ago, every technology vendor in the industry is espousing AI capabilities. There are innumerable potential use cases for AI in a typical life sciences organization, but many leaders lack a current perspective on the emerging field and the opportunities it creates.
For many leadership teams, getting a briefing from industry experts on current market research around AI in life sciences is a great tactic. In addition to establishing a common understanding among the team, the information stokes ideation around areas of potential AI focus within their enterprise.
3. Sponsor an AI Strategy Steering Committee
Artificial intelligence is expected to become one of the most prominent drivers of modernized businesses, including those in life sciences. Whether organizations have a plan for AI or not, AI-enabled tools will make their way into the enterprise through vendor systems, office productivity tools, mobile applications, user-propagated tools, open-source offerings, security products, and more. The question for leadership teams is not “whether” it will happen but rather “how” it will happen.
Though it is possible to manage AI diffusion through tactical decisions in individual functions, many life sciences companies will benefit from having a governance model empaneled to help oversee and steer their company’s progression and policies. In addition to avoiding unnecessary proliferation of AI products, steering committees can be particularly helpful in crafting the corporate strategy for AI, prioritizing protections (e.g., information controls), developing shared services, aligning hiring plans, ensuring training, and facilitating the cross-functional collaborations needed for more transformative applications of AI (e.g., enterprise workflow automation, predictive analytics).
4. Build an AI Strategic Plan
Speaking of strategy, AI is not simply a new shiny tool – it represents a strategic business opportunity. AI capabilities, such as generative AI, will increasingly permeate how we complete basic everyday tasks like taking notes, writing emails, developing meeting summaries, and searching the web. But for many organizations, the impact of AI will be strengthened due to its ability to transform business processes (e.g., efficiencies from automation), enhance performance (e.g., ensuring quality and safety through real-time oversight), and accelerate product innovation (e.g., discovery, design, and development process functions).
Given that the creation of these capabilities requires investment, how do leaders know where to place their bets? Strategic planning programs can help leaders parse the considerations, providing a framework for aligning around goals and imperatives for business growth. When these strategic planning programs include market research and internal capability assessments, leaders can more effectively tailor their AI strategies to specific opportunities that will translate into market differentiation and measurable value creation.
5. Use What You Already Know
Does your organization use Microsoft products? Life sciences organizations usually have a broad base of Microsoft users within their enterprises. Microsoft technologies such as Word, Outlook, and Teams are staple applications for knowledge workers, and developers have increasingly turned to Microsoft’s Azure environment for the development and deployment of cloud-based solutions.
Microsoft’s current strategy is to infuse artificial intelligence across their entire product portfolio. With some thoughtful planning, industry leaders can empower their employees to automate certain day-to-day tasks and streamline processes through Microsoft products that they are already using. Microsoft’s Copilot capabilities, for example, provide some AI natural language processing and deep learning within Office applications. And Azure’s AI services connect to Microsoft’s authentication environment, offering a technical path for securing the use of emerging AI solutions and their data.
6. Sponsor an AI Prototyping Workshop
Using out-of-the-box AI capabilities is a great way to get started, but they don’t necessarily lead to out-of-the-box thinking. To find more transformative AI opportunities for improvement and growth, leaders need to foster creativity and innovative thinking about new ways of working. Some of the most impressive opportunities for AI emerge when industry and company experts solicit ideas around:
- What are the hardest parts of what we do?
- What aspects of the work are so routine that a machine might be able to do it?
- What information and insights would help improve speed, quality, or satisfaction?
- How could we get a ten-fold improvement over where we are today?
- If we were to totally re-write how we do this today, what would it look like?
One way of exploring the “world of the possible” is by hosting an artificial intelligence prototyping workshop. These executive-sponsored events, which often run for two days, provide a hands-on forum for employees to brainstorm how AI models can be applied to business challenges. Teams of internal experts are paired with external AI consultants and compete on developing a working demonstration of AI’s applicability to their business. In addition to gaining experience with AI, employees learn how to identify AI improvement opportunities.
7. Create an AI Sandbox for Open Source
Besides specific workshops, it is also valuable to create dedicated spaces for employees to be able to work with artificial intelligence. The open-source community is alive with AI-related capabilities – models, code libraries, and more (e.g., AlphaFold) – that can be used to learn about and explore AI-powered improvements. However, due to the public nature of these assets, risks related to security, quality, service level disruption, and compliance need to be contained. And since many AI solutions need access to company data in some form, many organizations prefer to host these technologies in environments where they have more direct control of leveraged data sets.
Cloud and “virtual machine” technologies offer great mechanisms for standing up short-term “sandboxes” for exploring new AI projects. Most major cloud hosting companies have semi-automated procedures for quickly standing up and configuring new environments. Code access is restricted to information explicitly provided in the sandbox, and external communications can be configured to prevent disclosure of company information to third parties. Once the exploration is done, shutting the environment down is almost as easy as pressing the delete key.
8. Launch a Pilot Project
As life sciences organizations explore the world of possibilities with artificial intelligence, eventually at least one idea will surface that holds promise. Though AI opportunities may look like IT projects, the greatest impact will be felt when leaders treat AI investments as more comprehensive business initiatives.
What does that mean? Consider the following criteria in spinning up an AI pilot:
- Strategy & Objectives: are the goals clear and measurable? Does everyone understand how the goals align to the company’s strategy?
- Scope & Resourcing: is the project sized for success (i.e., specific, focused, impactful)? Do we have the right business and technical experts engaged?
- Use Case and Success Criteria: is the workflow (both “as-is” and “to-be”) well defined? Does everyone agree on what a great outcome looks like, and what measurable value is being generated? Do we know when we are done with this project?
- Architecture & Data Strategy: have we considered how the design of this AI capability might fit within our enterprise architecture (e.g., integration, authentication, authorization)? Do we know how this solution would scale up over time? Do we have a plan for providing the AI high-quality data? Are we clear on how to address privacy, regulatory, or data use constraints?
- Governance & Communication: are executive sponsorship and decision-making accountabilities clear? Are all stakeholders on board and willing to implement changes if the project is successful? Have we agreed on how, when, and where to share information and updates about the project?
- Quality & Compliance: do we know how we will assess the quality of the work we produce, the data we provide, and the results generated by AI? Are testing, validation, and user acceptance requirements and procedures clear?
- Deployment & Change Management: do we know how we intend to deploy, transition, and train? What are our plans for documentation, maintenance, support, and continuous improvement?
- Risks & Contingencies: have we agreed on how to assess and manage risk? Do we know what to do if the solution does not work as intended? What, if any, ethical concerns need to be managed?
Of course, you don’t need the answer to all those questions to get started with a pilot. But ensuring leaders and pilot teams are thinking holistically about the program will help to ensure that successful pilots can be converted to successful new business capabilities.
If you’d like to learn more about AI technologies in life sciences, consider reading my white paper. It provides an overview of some of the key trends life sciences leaders need to be aware of when considering AI investments. And reach out to me directly if you’d like to talk more about any of these concepts, I’m always curious to learn how others are thinking about this exciting space.
