A growing gap has emerged between how life sciences and pharma companies view the potential of agentic AI and how far those technologies have progressed in real-world use, according to new data released by automation software provider Camunda.

The company’s 2026 State of Agentic Orchestration and Automation report found that while experimentation with AI agents is widespread across the sector, only a small proportion of use cases are reaching production. The findings are based on responses from life sciences and pharma organisations and highlight persistent concerns around risk, governance and operational maturity.

According to the report, 68% of organisations in the sector said there is a disconnect between their agentic AI vision and current reality. Although 73% reported using AI agents in some form, just 11% of agentic AI use cases had reached production over the past year. Nearly half of respondents said poorly controlled agentic AI could worsen existing process and automation challenges.

Trust emerged as a central barrier to wider deployment. The majority of respondents cited concerns around governance and transparency, with 87% worried about business risk when AI is deployed without appropriate IT controls. A further 79% said they lack sufficient visibility into how AI is being used across processes, while 66% pointed to compliance concerns as a limiting factor.

As a result, most AI agents currently in use remain confined to low-risk applications. The report found that 76% of organisations said their agents are primarily chatbots or assistants used for summarisation or question answering, rather than being embedded in mission-critical workflows. Nearly half said their agents operate in silos rather than being integrated into end-to-end processes.

Kurt Petersen, senior vice president customer success at Camunda, said: “The promise of agentic AI is undeniable, but trust remains the key barrier to adoption.”

He added: “Right now, exercising caution with agentic AI means many organizations can’t move beyond pilots or isolated use cases. Without clear guardrails and visibility, agents will stay at the edge of the business.”

Despite these challenges, automation more broadly continues to deliver measurable impact across the sector. The report found that 92% of life sciences and pharma organisations reported business growth linked to process automation over the past year. On average, companies said they have automated just over half of their processes, with expectations this could rise to nearly two-thirds.

Investment levels are also increasing. Around 77% of respondents said they plan to raise automation spending, with budgets expected to grow by an average of 18% over the next two years. However, this expansion is happening alongside increasing technical complexity. Seven in ten organisations said the number and diversity of endpoints involved in their processes is growing rapidly, creating additional orchestration and oversight challenges.

As systems become more distributed, 82% of respondents said they need improved tools to manage interactions between processes, systems and AI-driven components in order to realise the full value of their investments.

The report positions agentic orchestration as a potential solution, combining structured process automation with AI-driven reasoning. However, most organisations acknowledge they are not yet ready to implement this approach at scale. While 85% said AI needs to be orchestrated across business processes to maximise value, 80% said they have not reached sufficient process maturity to do so effectively.

Petersen said: “Agentic orchestration, not standalone agents, is the key to closing the AI vision-reality gap.”

He added that blending deterministic process controls with dynamic AI-driven decision-making could allow organisations to deploy agents inside governed workflows, rather than as isolated tools.

While the report is vendor-led and reflects Camunda’s positioning around orchestration technology, the findings underline broader industry challenges facing pharma and life sciences companies as they move from AI experimentation toward operational deployment in regulated environments.