How to Plan Your AI Budget Now To Succeed in 2025

In 2025, AI will be central to business strategy, especially in life sciences, with significant investments expected. Over time, AI is anticipated to reduce the cycle length and failure rate in scientific research and drug discovery, lowering the cost per approved drug. Reports indicate that the life sciences market size for AI investment could reach nearly $10 billion by 2032, and 40 percent of pharma companies are already including anticipated savings from generative AI in their 2024 budgets.

However, rising costs are a primary threat to AI's success. Generative AI (GenAI) is rapidly being integrated into the life science value chain, but its costs and ROI remain largely unknown. Surveys show that while most life science organizations have GenAI use cases in production or pilot, more than half abandon their efforts due to cost-related missteps.

Because the AI space is evolving quickly, companies must budget correctly and use investments strategically. Here’s how to plan your AI budget for success in the coming year.

Spend Money to Help Scientists Increase the Value of Data

AI investments should make scientists' work easier by supporting the Lab-in-a-Loop lifecycle, which integrates physical (wet lab) and simulated (dry lab) experiments. Key considerations include:

• ​Scientifically-smart technologies: Provide scientists with access to the data they need, using science-first tools to reduce costs by capturing only relevant data.
• ​FAIR data & processes: Enable scientists to gather, understand, and use data without heavy manual tagging. FAIR (Findable, Accessible, Interoperable, Reusable) data and processes increase transparency and reproducibility.
• ​Future-proof technologies: Use tools that scale with your organization and industry, allowing flexibility and seamless collaboration across departments.
• ​Low-code or no-code technologies: Empower scientists to reduce reliance on data science and IT, expanding the value of AI/ML.

Put Your AI Dollars Where They Matter Most

Efficient AI spending requires a clear strategy. Major cost drivers include data, compute, and people. Proper data architecture and infrastructure are essential—"garbage in, garbage out" applies. Maximizing storage, compute, security, GPUs, training, and storage costs is crucial, as is integrating data strategies with existing software investments.

Data is often the most volatile cost due to quality, availability, and governance challenges. Proprietary data, built over years of experiments, is expensive but differentiates companies. More data doesn't always mean better models, but using it effectively at scale is costly. The "Data-in-a-Loop" process—creating, using, and feeding data back into models—is expensive to support.

Maintenance is another significant cost. Buying data is just the start; modeling, cleansing, labeling, and aligning with ontologies are complex and time-consuming. Instrumentation challenges and maintaining on-premise or cloud solutions add IT overhead and increase R&D costs.

Early adopters in life sciences are outpacing others, with some deploying 11 or more use cases in production. These leaders are defining AI application pathways for accelerating drug discovery.

Leading organizations will prioritize high-return use cases aligned with future business goals. AI investments are often billed to IT, though benefits accrue to functional budgets. It's important to incentivize business unit leaders to invest in value-generating AI initiatives.

Tips to Setting Your AI Budget Effectively

• ​Understand pricing metrics and scaling: Analyze how AI product pricing structures scale with use, as vendors differ (e.g., tokens vs. character counts).
• ​Standardize vendor assessment: Use a consistent approach to model annualized costs, including fine-tuning, configuration, and retraining.
• ​Minimize unbudgeted costs: Negotiate scalability and require transparency about hidden costs, such as overages in prepaid credits.
• ​Integrate key metrics early: Identify and track metrics from the ideation phase to measure impact and ROI post-deployment. Determining tangible outcomes and financial impact is a top challenge for many leaders.
• ​Consider cost savings: New AI tech can replace existing technologies and cut costs. Prioritize use cases that free up scientists' time, such as assistive AI for repetitive tasks, to reduce admin costs and errors. This enables scientists to focus on training and improving models, supporting the Data-in-a-Loop workflow.

Scientific Lens on People, Data, and Models

As AI in life sciences evolves, CIOs must balance investments with the ability to safely and effectively implement and scale technologies. While AI offers tremendous promise, organizations are built on people. Technology should not overburden scientists; instead, it should enable FAIR data and processes, ensuring that scientific work remains central. The role of the scientist must be prioritized in developing AI policies and technologies.

References

1. 1.AI In Life Sciences Market Size, Companies and Future Analysis: https://www.towardshealthcare.com/insights/ai-in-life-sciences-market
2. 2.How to Successfully Scale Generative AI in Pharma: https://www.bain.com/insights/how-to-successfully-scale-generative-ai-in-pharma/
3. 3.Q24 LLMs and Generative AI: Life Science Manufacturer Perspective: https://www.gartner.com/document-reader/document/5638991?ref=sendres_email&refval=79002868
4. 4.Q24 LLMs and Generative AI: Life Science Manufacturer Perspective: https://www.gartner.com/document-reader/document/5638991?ref=sendres_email&refval=79002868
5. 5.Q24 LLMs and Generative AI: Life Science Manufacturer Perspective: https://www.gartner.com/document-reader/document/5638991?ref=sendres_email&refval=79002868