Skip to main content

Transforming AI Potential into Meaningful Outcomes: A Strategic Action Plan for Life Sciences Companies

Foreword

While the Life Sciences industry has never lacked ambition, it often faces a lack of urgency. Every boardroom today is talking about GenAI and most companies are running "pilots" of "proof of concepts" (PoC’s). However, a pilot or PoC is not a strategy. A PoC that never scales is of limited value and does not extract the real value from AI.

In Switzerland, we have a unique advantage: global headquarters, scientific depth regulatory credibility and centres of innovation. But these assets will be immaterial in the absence of a clear strategy supporting targeted action. The real competitive edge in GenAI enablement will come from a defined, value-led, and business-enabled time bound action plan – showing measurable results in six months, scaling impact in twelve and embedding GenAI as an operating system of organisations in two years.

We need to re-engineer how we discover, develop and deliver products and experiences that matter to patients. It is about turning fragmented experiments into a disciplined portfolio of opportunities. And it is about refusing to let the "middleware permafrost" of outdated systems and operating models deter and delay what could be a transformative revolution.

Yet, while the AI-promise is clear, there are execution challenges. A recent MIT report revealed that 95% of GenAI pilots are failing to scale or deliver measurable value. Perhaps the problem is not the technology itself, but how enterprises integrate and operationalise it.

Turning Failure into Opportunities

Opportunity 1: From Isolated Pilots to Enterprise Breakthroughs

Aggregate learnings into an AI Value Heatmap and convert islands of innovation into enterprise lighthouse projects. Teams often conduct limited-scale experiments that do not extend beyond specific use cases. This ultimately hampers the potential for broader adoption and impact across the organisation.

Pharma should treat proofs of concept as a portfolio of opportunities, not as isolated experiments. Some will fail, but collectively they sharpen strategy – and the best can scale into enterprise-wide change.

– Nishant Sinha, Director, AI & Data, Deloitte

Opportunity 2: From Talent Gaps to a Future-Ready Workforce

Launch role-specific AI academies, embed copilots, and train leaders in AI fluency. While off-the-shelf solutions may appear impressive during demonstrations, they could fail to integrate smoothly with existing enterprise workflows, leading to frustration rather than delivering measurable outcomes.

The biggest risk is not that GenAI will replace people – it is that companies will fail to reskill their people to work with GenAI. Talent gaps are not barriers, they are the fastest route to build a workforce that is future ready and AI-fluent.

– Marc Beierschoder, Partner, AI & Data Lead, Deloitte

Opportunity 3: From Regulatory Uncertainty to Global Leadership

Collaborate and enable experimentation with partners and tech providers and prioritise areas to establish enterprise level AI governance.

Too many GenAI initiatives melt away when they hit the ‘middleware permafrost’ of outdated systems, siloed data and legacy governance. Executives must confront this head-on, or agentic AI will remain a sideshow instead of reshaping the enterprise.

– Antonio Russo, Partner, European AI Lead, Deloitte

Current landscape of AI integration in Life Sciences organisations across their life stages

From Deloitte’s engagement with Swiss Life Sciences companies, we have ascertained broadly three evolutionary stages where AI/GenAI enablement can potentially deliver impact and enable a frictionless transition:

Pre-commercial and first-product-launch planning. GenAI can ingest vast biomedical datasets and generate hypotheses at scale. By automating protocol authoring, literature reviews, and trial design, companies can cut documentation cycles by up to 30-40%.

As companies prepare to launch, AI can support contract management, customer-led packaging design, automated content tagging, and value proposition/dossier creation focused on regulatory compliance enablement. Dynamic GenAI-driven content creation ensures scientific accuracy while tailoring communications. Commercial teams already see 20% savings in content development costs with AI copilots.

With growth, AI can unlock efficiency in support functions such as chatbots for employee queries on HR policies or financial analytics to prevent synthetic fraud data generation to undertaking commercial gross-to-net analysis.

Focusing on AI initiatives that address the specific needs of an organisation is critical. It ensures that AI initiatives address the most pressing needs while aligning with organisational maturity and resource availability to drive maximum impact.

Value-led, Business-enabled Action plan

From our experience across a variety of AI projects with companies, in particular, emerging and next-generation Life Sciences companies, we have gathered insights that support the following three step approach towards a structured action plan:

Create a value heatmap of potential AI applications across the company’s value chain. Assess your landscape and look across your business processes/functions. This helps leaders to identify and prioritise the use cases with the most potential impact. This should be a 1-2 weeks exercise based on insights from key opinion leaders and process owners across the enterprise.

Analyse selected use case priorities in more detail by assessing both expected value and practical feasibility. This should be a 3-4 weeks assessment exercise together with the financial controllers to ensure the organisation focuses on use cases where technology, data, and organisational readiness align with feasibility and impact.

Finally, implement a minimum viable pilot (MVP) in a controlled environment. The goal is to demonstrate tangible value within 8-10 weeks, while laying the foundation for scaling.

This action plan can be enabled for Life Sciences organisations at different stages of the life cycle, such as a biotech company at the pre-revenue stage or a MedTech firm approaching commercialisation or a scaling mid-cap.

With Deloitte experience anchored in the lean approach and continuous value realisation, companies can enhance their success on the AI Enablement journey. This will pave the way for future advancements in AI adoption to drive sustainable growth and a competitive advantage without mobilising huge amounts of resources.

Moving forward with the right AI partner

The path is clear for Swiss Life Sciences companies who aim to:

  • accelerate research and development through automation and advanced intelligence,
  • streamline commercialisation with more efficient and effective launch processes
  • scale operations with cost optimisation by empowering essential functions with AI capabilities.

Companies will need to move beyond unplanned AI experimentations.

They will need to embrace a structured and lean roadmap and align AI initiatives with their lifecycle stage and learnings. They will also need to select the appropriate model of sustainable AI that aligns with their goals and aspirations to deliver value quickly. Only then will companies thrive in the age of AI.

Outlook for the future

At Deloitte we recognise that AI is not just a trend, but a powerful catalyst for growth. Our mission is to assist organisations in transforming pilot projects into tangible impact which leads to sustained growth.

The next wave of competitive advantages for Swiss Life Sciences companies will not come from isolated pilots, but from AI that is embedded, scalable, and sustainable. They have to act now: If they can’t show results in six months, they will have lost momentum. By twelve months, competitors will have scaled lighthouse projects to drive enterprise impact.

"Those who take a patient outcome and business-value based, strategic approach today will shape the transformation and impact of the industry in the near future."

– Hannah Hayward, Partner, CIO Programme Chair, Deloitte

Contributors

We thank Gurneesh Cheema, Jasa Andrensek and Ruta Verbickaite for their contributions to the article.

Did you find this useful?

Thanks for your feedback