SORA 2.5: What Has Changed for UK and EU Operators

SORA has evolved through successive editions, and version 2.5 represents the latest significant revision of the JARUS methodology. For UK and EU operators, understanding what changed — and what did not — helps you build a current, defensible safety case. This guide summarises the nature of the 2.5 revision and how to approach it.

The core logic is unchanged

First, the reassuring part: the fundamental architecture of SORA survives 2.5 intact. You still describe a Concept of Operations, assess Ground Risk Class and Air Risk Class, apply mitigations, derive a SAIL, and demonstrate Operational Safety Objectives. If you understand the GRC/ARC/SAIL chain, you already understand the backbone of 2.5.

A refined semantic model

A major theme of 2.5 is clarity. Earlier editions left some terms open to interpretation, which led to inconsistent application between operators and authorities. Version 2.5 tightens the semantic model — the precise definitions of the concepts used — so that the same words mean the same thing across the industry. This reduces ambiguity in how operations, volumes and buffers are described.

Terminology refinements

Alongside the semantic model, 2.5 refines terminology so that the language of the methodology is more consistent and precise. Operators moving from an earlier edition will recognise most concepts but should check that they are using current terms, because assessors expect the up-to-date vocabulary in a submission.

Clarified ground-risk and air-risk processes

The 2.5 revision clarifies how the ground-risk and air-risk parts of the assessment are worked through. The intent is to make the steps more deterministic and less open to differing interpretation — for example, in how the operational volume and ground-risk buffer are established, and how the air-risk environment is characterised. The underlying principles (size and population for GRC; airspace and traffic for ARC) remain, but the process around them is more clearly defined.

Mitigations and robustness

The treatment of mitigations and their robustness is refined to be more consistent. The distinction between integrity (the measure works) and assurance (you can prove it) remains central, and 2.5 aims to make the expectations for each robustness level clearer so operators can judge what evidence is needed.

What this means for UK operators

The UK CAA aligns its guidance with the JARUS methodology, and references the relevant edition in CAP 722. The practical implications for UK operators are:

• Use the current edition. Build your SORA against the version the CAA currently references, not an older one you may have used before.
• Update your terminology. Ensure your submission uses the current terms and definitions.
• Re-check your tables. Where GRC/ARC values or OSO robustness expectations have been refined, confirm you are reading from the current tables.
• Don't assume older examples transfer. A worked example from an earlier edition is illustrative only; the detail may have moved.

For operators working across the UK and EU

Because both the UK and EU draw on the same JARUS methodology, a 2.5-aligned safety case shares a common conceptual foundation across both. This does not mean an authorisation in one is automatically valid in the other — the regulatory frameworks remain separate — but the underlying risk logic is broadly recognisable, which eases the work for operators active in both jurisdictions.

Practical advice

If you have an existing SORA built against an earlier edition, treat 2.5 as a prompt to review rather than rebuild: confirm your terminology, re-check your GRC/ARC derivations against current tables, and make sure your mitigation and OSO evidence still maps to the current robustness expectations. Always confirm the precise current requirements with the CAA's published guidance, because the detail of any edition can be superseded.

SORA 2.5 is best understood as a maturing of a methodology that already worked — sharper definitions, clearer processes and more consistent language, built on the same proven GRC/ARC/SAIL foundation.