Life sciences and pharmaceutical organisations report a growing gap between their ambitions for agentic AI and real-world implementation, according to new industry research. While adoption of AI agents is widespread, few use cases have progressed into full production environments.
Agentic AI adoption outpaces production readiness
The findings come from Camunda’s 2026 State of Agentic Orchestration and Automation report, which examined AI and automation maturity across regulated industries. Among life sciences and pharma respondents, 73% said they currently use AI agents.
However, only 11% of agentic AI use cases reached production over the past year. In addition, 68% of organisations acknowledged a gap between their agentic AI vision and operational reality.
Trust, transparency and compliance remain key barriers
Importantly, the report highlights persistent concerns around governance and risk management. Almost nine in ten respondents cited business risk when AI is deployed without sufficient IT controls.
In addition, 79% reported concerns about transparency into AI usage, while 66% identified compliance risks associated with deploying AI agents in regulated environments.
AI agents largely limited to non-critical roles
As a result, most AI agents remain confined to low-risk use cases. Around 76% of organisations said their AI agents are primarily used as chatbots or assistants rather than supporting mission-critical processes.
Meanwhile, 49% reported that AI agents operate in silos and are not integrated into end-to-end business workflows.
Automation investment grows despite rising complexity
The report also found that 92% of life sciences and pharma organisations experienced business growth linked to automation during the past year. On average, 51% of processes are currently automated, with respondents estimating this could rise to 64%.
Looking ahead, 77% plan to increase automation investment, with budgets expected to grow by an average of 18% over the next two years.
Process orchestration seen as critical next step
At the same time, 70% of organisations reported rapid growth in the number and diversity of system endpoints involved in business processes. Consequently, 82% said they need improved tools to manage interactions across increasingly complex environments.
The report found strong consensus that AI must be orchestrated within governed processes. However, 80% of respondents said they have not yet achieved the level of process maturity required to implement agentic orchestration effectively.
Further details are available on the Camunda website. Related coverage can be found in the Life Sciences Global News Digital Health & Health Tech section.
