Issue 10 is being written right now. What you fix this year is what it will audit.

The public consultation on revising the BRCGS Global Standard Food Safety closed on 15 February 2026. More than 570 food safety professionals applied to contribute to the technical working groups. Issue 10 is being drafted now, and nobody outside those rooms knows what it will say.

Which makes this a useful year. Issue 9 is still the standard your audit is conducted against, across more than 24,500 certificated sites in 137 countries, and the clauses sites lose points on are already the ones about evidence rather than intent. Fix those and you are ready for whatever the numbering changes to.

What Issue 9 already asks for, and where sites come unstuck

Clause 4.11.8 is the environmental monitoring clause. Issue 9 moved it on from having a programme to being able to defend one: documented sampling procedures, action limits that distinguish a food safety objective from the contaminant levels you actually find in a production area, and trending of results. A folder of certificates of analysis is not a trend. Neither is a spreadsheet nobody has plotted.

Section 8 consolidates the requirements for high risk, high care and ambient high care zones, with the zone definitions in Appendix 2. Clause 4.3.1 designates the zones, 4.3.2 and 8.1.1 require them identified on site maps, and 8.1.2 to 8.1.4 deal with physical segregation and restrictions on the movement of people, materials and equipment. If your zone map and your actual traffic pattern have drifted apart since the last line change, that is a finding waiting to happen.

Three more Issue 9 clauses are worth reading again with an auditor's eye. 8.2.3 wants documented procedures for removeable walls in high risk and high care areas, tight fitting, opened only by trained staff, cleaning completed before production restarts. 8.5.4 requires designated cleaning in place systems or controls for different production areas. And 8.7.1 asks for validation of cleaning effectiveness, not merely testing of it, which is a different and harder claim to make.

Then the two clauses that decide how a bad week is judged. Clause 3.7 requires root cause analysis inside the corrective action procedure, with documented preventive action. Clause 3.11 requires a significant food safety incident to be notified to your certification body within three working days, with full information including root cause analysis and a preventive action plan within 21 calendar days. Twenty one days is not long to reconstruct what a room looked like three weeks ago.

The statutory floor underneath the certificate

BRCGS is a customer requirement. The law is separate and sits underneath it. Article 5 of Regulation (EC) No 852/2004, now assimilated in UK law, requires a permanent procedure based on the HACCP principles, and Article 5(2)(g) requires documents and records commensurate with the nature and size of the business. Section 21 of the Food Safety Act 1990 gives you the due diligence defence, which is only ever as good as the records that support it.

For ready to eat food that may pose a Listeria monocytogenes risk, Article 5 of Regulation (EC) No 2073/2005 goes further and requires the food business operator to sample the processing areas and equipment for Listeria monocytogenes as part of its sampling scheme, with ISO 18593 as the reference method for environmental sampling. Annex I sets the food safety criteria: not detected in 25g for products intended for infants or special medical purposes, and 100 cfu/g through shelf life for other ready to eat products, or not detected in 25g before the product leaves your control where growth is possible.

Environmental monitoring is not a best practice add on. For a ready to eat manufacturer it is a legal requirement with a named reference method.

The pressure is not coming from the council

If you were waiting for local authority inspection to be the forcing function, the published numbers say otherwise. In its June 2026 performance update, reporting data as at October 2025, the Food Standards Agency recorded 72,000 overdue food hygiene interventions in England, down from 95,000 a year earlier but still well above the pre pandemic baseline of 46,000. Over 39,500 newly registered businesses were awaiting their first inspection, out of 585,000 registered premises. Category A and B premises have largely recovered to pre pandemic intervention rates. Categories D and E have not.

So fewer visits, but harder ones. Formal enforcement for food hygiene reached 6,199 cases in 2024/25, a 30 per cent increase on the pre pandemic 4,784, and food standards enforcement rose 47 per cent. Seven local authorities in England are in escalation. Demand for officers is growing faster than the supply of them, which is the FSA's own analysis, not ours.

The real pressure comes from your customer's audit programme and from what happens when a positive result lands.

What a positive result costs, in the only currency that matters

UKHSA recorded 181 confirmed listeriosis cases in England and Wales in 2025, an 8.4 per cent increase on the previous five year median. Among the 141 cases not associated with pregnancy, 28 people died. Of the 40 pregnancy associated cases, just under a third ended in stillbirth or miscarriage. Four outbreaks were investigated, with sources including smoked fish and prepacked sandwiches.

We are not going to put a pound figure on a UK recall, because there is no properly sourced published one. Every number in circulation traces back to vendor marketing. The FSA's own evaluation of withdrawal and recall effectiveness explicitly excluded economic impact because attribution was too difficult. What the FSA has published is the societal cost of foodborne disease in the UK, £9 billion a year on 2018 data, and that figure is doing enough work on its own.

Why one verification tool is never enough

This is uncomfortable but it is published. A systematic review and meta analysis in the Journal of Food Protection in 2022 found the relationship between ATP bioluminescence readings and microbial assessments in food establishments to be weak, with a random effects correlation of about 0.30. The authors' conclusion was not that ATP is useless. It was that a food business cannot depend on one method.

A 2025 field study in the same journal put that to work in an avocado packing plant. ATP readings showed no significant reduction after cleaning and sanitising. No surface achieved the expected three log reduction. Listeria species were detected on 14 per cent of food contact surfaces before cleaning and 13 per cent after. The verification tools contradicted each other: some surfaces looked clean and read clean on microbial indicators while ATP showed no change, others improved visually with nothing to corroborate it.

Any technical manager who has argued with a swab result already knows this. The published evidence just makes it quotable.

Where a spatial record fits

A 3D capture, meaning a photorealistic and measurable three dimensional reconstruction of a room, does one thing well: it fixes what a space looked like on a given date, in a form you can navigate, measure and annotate afterwards.

For a hygiene programme that means the sampling points sit on the room rather than in a list, so a new technical manager can see where site 47 actually is. It means the zone map and the room agree, because the room is the map. It means that when a line is rebuilt, the before and after are directly comparable, and the movement route you claimed in the segregation procedure can be shown rather than described. And when clause 3.11 gives you 21 days to produce a root cause analysis, the condition of the room three weeks ago is still available to look at.

On the optical science, be sceptical of anyone quoting factory performance. The 2026 literature is genuinely promising and genuinely early. A study in Optics Express in January 2026 imaged single and dual species Listeria monocytogenes biofilms on stainless steel and PVC coupons with 91 and 89 per cent accuracy using handheld fluorescence imaging. Work in Analytica Chimica Acta the same month identified pathogens in mixed biofilms at 90 to 99 per cent accuracy using hyperspectral imaging and machine learning. Both are coupon and in vitro studies. Every recent review of hyperspectral imaging in food names the gap between laboratory conditions and a production line as the principal unresolved obstacle, and so do we.

What to do before Issue 10 lands

Plot your environmental monitoring results and see whether you have a trend or a pile. Check your zone map against the traffic your site actually runs. Ask whether your cleaning validation would satisfy 8.7.1 or only your own habit. And decide now where the dated record of each high care room lives, because that is the part you cannot reconstruct after the event.

Issue 10 will be written by people who audit food factories for a living. It will ask for better evidence, because every issue of every standard does. The sites that pass it comfortably will be the ones that started this year. Our related pieces cover the blind spot between swabs, what hyperspectral imaging actually sees and how a proving phase works on a live line.

This article is general information about UK food law and third party certification and is not legal or technical advice. Food business operators should take their own advice on their specific obligations, and should work from the current published text of the BRCGS standard.