CKO-025Human Oversight & AutonomyStrong evidenceHigh risk

Are autonomous end-to-end evidence synthesis systems trustworthy?

Current evidence does not support fully autonomous end-to-end evidence synthesis systems for decision-grade evidence production.

In more detail

End-to-end systems attempt to automate searching, screening, extraction and synthesis within a single workflow. Errors can accumulate across stages, making outputs difficult to verify. Current RAISE guidance considers such systems unsuitable for formal evidence synthesis without substantial human oversight.

Why it matters

Decision-makers may mistakenly assume automated outputs are equivalent to systematic reviews.

Decision rule

Do not treat autonomous end-to-end outputs as decision-grade evidence.

Common misconception

  • “If each component works reasonably well, the entire system must be trustworthy.”

At a glance

Evidence strength
Strong
Risk category
High
Trust impact
High

Related concepts

Agentic AI Human Oversight Summarisation vs Synthesis
Key takeaway

Autonomous evidence synthesis remains a future aspiration rather than a current reality.

More on Human Oversight & Autonomy