Food Defense: what it is, what it requires and how to train
The hazard your HACCP study will never find, because the process does not cause it: someone who wants to cause it does. And it is rarely a stranger climbing the fence. It is usually someone who already has access, a key and a shift pattern.
Food defense is the protection of food from intentional contamination or adulteration committed with the intent to cause harm, whether to public health or to the business. It is assessed through a threat methodology known as TACCP (Threat Assessment and Critical Control Points), which identifies who might attack, through which point, with what motivation and to what consequence, and determines the measures that make such an attack hard or detectable. It is required by every certification scheme recognised by the Global Food Safety Initiative (GFSI) and, for sites exporting to the United States, by the FDA intentional adulteration rule.
In short
- Motivation
- To cause harm (unlike food fraud)
- Methodology
- TACCP — threat assessment
- Required by
- All GFSI-recognised schemes
- Reference guidance
- PAS 96:2017 (BSI, free to download)
- US regulation
- 21 CFR part 121 — intentional adulteration
- Mandatory since
- July 2019 for large businesses; 2020 and 2021 for the rest
- FSSC 22000 V7
- Additional requirement 2.5.3, food defense
- BRCGS Food v9
- Clause 4.2, site security and food defence
- IFS Food v8
- Chapter 6, food defense
- SQF Ed. 9
- Clause 2.7, food defense and food fraud
This guide is written for whoever has to learn or teach Food Defense. If what you need is for us to implement the system at your company, that work is done by Beira Consultores.
See consulting at beira.com.mxFood defense and food fraud are not the same thing
Both methodologies protect against a deliberate act, and that is where the similarity ends. In food defense (TACCP) the motivation is to cause harm: the attacker wants the effect to happen and, often, to be noticed. In food fraud (VACCP) the motivation is economic: someone makes money if the product passes for what it is not, and an outbreak is the last thing they want, because it draws exactly the attention that ruins the business.
The consequence is operational, not terminological. A camera does nothing against the economic attacker, whose adulteration is designed not to show up in analysis; access control, custody of chemicals and supervision of the step where one person is left alone with open product do work against the attacker who wants to cause harm. That is how a site can hold a flawless fraud vulnerability assessment and still leave the receiving tank without a lock.
Schemes ask for two assessments and two plans. Presenting them as one undifferentiated document is a common finding, and it is easy to see where it comes from: the team is the same and so is part of the background information. What cannot be shared is the scoring, because one measures access and intent to harm while the other measures economic incentive.
What your scheme requires, exactly
Since version 7 of the GFSI benchmarking requirements, in force from 1 January 2018, every recognised scheme asks for two food defense deliverables. Not one: two, and the second is the one that ends up half done.
- A documented threat assessment covering the scope of the certification.
- A defense plan derived from that assessment, with measures assigned, verifiable and reviewed.
- FSSC 22000 requires it in additional requirement 2.5.3, separate from 2.5.4 on food fraud mitigation.
- BRCGS Food Safety Issue 9 covers it in clause 4.2, together with site security.
- IFS Food version 8 devotes a full chapter to it, chapter 6, with team, scope and review documented.
- SQF edition 9 covers it in clause 2.7, in the same place as food fraud.
- ISO/TS 22002-1 includes it among the prerequisite programmes, in clause 18.
- ISO 22000:2018 on its own does not require it explicitly: it arrives with the GFSI scheme built on top of it.
The FDA rule reaches sites that have never set foot in the United States
The FDA intentional adulteration rule (21 CFR part 121) applies to facilities required to register with the FDA, and that includes foreign sites exporting food to that market. A plant in Mexico can fall under it without a single operation on US soil, and NOM-251 does not mention it anywhere.
What it asks for is a written food defense plan with five pieces: the vulnerability assessment that identifies the actionable process steps, the mitigation strategies for each one, the monitoring procedures, the corrective actions and the verification. The plan is reanalysed at least every three years, and sooner if something relevant changes. The compliance dates have passed: July 2019 for large businesses, 2020 for small ones and 2021 for very small ones, which are also left with modified requirements.
The trap is assuming the certification plan already covers the rule. It does not: the FDA reasons by actionable process steps (with the key activity types as a recognised shortcut: bulk liquid receiving and loading, liquid storage and handling, secondary ingredient handling, and mixing and similar activities), while TACCP reasons by threat and by actor. The rule also names two roles the scheme does not require: the food defense qualified individual, who needs specific training to prepare the plan, and the personnel assigned to actionable process steps, who need documented awareness training.
Why the attacker is usually on the inside
PAS 96, the British reference guidance for TACCP, sorts threats into a handful of families: malicious contamination, extortion, espionage, counterfeiting, cyber crime and economically motivated adulteration. Of all of them, the one that has stopped the most plants is malicious contamination committed by someone in the house or by a contractor with legitimate access.
The logic is uncomfortable: the insider knows where there is no supervision, knows the shift in which a step is left unattended, and does not need to force any entrance. A plan that only lists fences, cameras and visitor control is answering the perimeter attack, which is the least likely one, and leaves open the one that actually happens.
The controls that change something are duller and cheaper: zone access by what each area holds instead of handing out master keys, never leave one person alone with open product at the critical steps, keep cleaning chemicals and allergens under custody, lock tank and silo hatches, control contractors in and out, revoke access the same day someone leaves, and keep a channel through which people can raise a concern without exposing themselves.
How to train in food defense
Training in this subject swings to one of two extremes, and both leave the learner unable to do the work. One is the site security course, which talks about locks, cameras and patrols and never reaches the process. The other is the compliance course, which recites which clause requires it and teaches no assessment at all. What is needed sits in between: identifying threats with explicit criteria, deciding which process steps are the vulnerable ones, and walking out with the defense plan written.
Nor is the audience only food safety. Access control is decided by the plant, hiring and terminations by HR, and a response to an extortion attempt by senior management. Training that stops at the quality team produces a plan nobody outside quality can run, which is exactly what the auditor finds when they start asking on the floor.
And if your site exports to the United States, training stops being optional at one specific point: the intentional adulteration rule requires the plan to be prepared by a qualified individual with specific training, and the personnel assigned to actionable process steps to receive awareness training that is kept on record.
Who needs training in Food Defense
Food safety or system manager
Build the full threat assessment and defend to the auditor why a process step was left out, not just hand over a signed plan.
Plant or operations manager
Decide which measures are workable without stopping production: access zoning, custody of chemicals, supervision of bulk receiving.
Food defense qualified individual
Prepare, verify and reanalyse the plan under 21 CFR part 121 at a site that exports to the United States, where the role requires specific training.
HR and site security
Cover the most likely and best documented scenario: the insider. Hiring, terminations, visitors, contractors and a way for people to report.
Online Food Defense courses
Defensa Alimentaria (TACCP) con PAS 96:2026 y el método de la FDA
Aprendes las dos metodologías vigentes y sales con tu plan de defensa escrito y firmado
Frequently asked questions about Food Defense
What is food defense?
Food defense is the protection of food from intentional contamination or adulteration committed to cause harm, either to the health of whoever consumes it or to the business. It covers insider sabotage, malicious contamination, extortion, espionage and cyber attacks against the systems that control the process. It differs from food fraud in intent: here the aim is to cause harm, not to make money.
What is the difference between TACCP and VACCP?
TACCP assesses food defense threats and asks who would want to cause harm, where they would get in and what stops them. VACCP assesses food fraud vulnerability and asks who profits from adulterating an input and by how much. The difference is not vocabulary: the attacker seeking harm wants the effect to be noticed, and the one seeking money needs precisely the opposite, so the controls that stop one do not stop the other.
Does my certification require a food defense plan?
Yes, if your scheme is GFSI-recognised. Every recognised scheme requires a documented threat assessment and a defense plan derived from it: additional requirement 2.5.3 of FSSC 22000, clause 4.2 of BRCGS Food Safety Issue 9, chapter 6 of IFS Food version 8 and clause 2.7 of SQF edition 9. ISO 22000:2018 on its own does not require it explicitly; the requirement arrives with the scheme built on top of it.
Do I have to comply with the FDA rule if I export to the United States?
If your facility is required to register with the FDA, yes, even if it sits outside the United States. The intentional adulteration rule (21 CFR part 121) requires a written food defense plan with a vulnerability assessment, mitigation strategies, monitoring, corrective actions and verification. There are exemptions and modified requirements, above all for very small businesses and for certain holding and packing operations, so the first step is to establish the scope rather than assume it.
How often is the food defense plan reviewed?
The FDA rule requires reanalysis at least every three years, and sooner if something that affects the assessment changes. Certification schemes are stricter in practice: they ask for review at least annually and whenever the process, the facility or the threat environment changes. A site covered by both follows the shorter interval, which is always the scheme's.
Do cameras count as a food defense measure?
A camera nobody watches is not a mitigation strategy: it is a recording for after the incident. To count as a control, the measure has to genuinely reduce the chance of the attack at that process step or make it detectable, and it has to have monitoring, corrective action and verification assigned to it. A lock on a tank hatch, a key under control and the rule of never leaving anyone alone with open product are usually worth more than closed-circuit television.