
The examinee no longer picks the examiner. FSSAI's faceless audit allocation, effective today, is the first real brick of Food Safety 2.0. The next one must be faceless sampling.
By Kuldeep Sharma, Chief Editor, DairyNews7x7
For eight years, an Indian food business could choose who would audit it, and then pay them directly. From today, it cannot.
With effect from 19 November 2026, the FSSAI faceless audit mechanism allocates recognised third-party auditing agencies through the Audit Management System (AMS), integrated with FoSCoS. The food business raises an audit request; the system, not the business, assigns the agency. Few will notice this reform. Few headlines will carry it. It still matters more to the credibility of food safety in India than the next hundred adulteration raids.
The reason is simple. In any audit system, the examinee should not pick the examiner.
What the FSSAI faceless audit mechanism changes
According to the FSSAI notification on the AMS portal, FBOs can no longer choose an agency manually. The new process works like this:
- The FBO raises an audit request on FoSCoS.
- The system allocates a recognised agency, which accepts or rejects the assignment on AMS. A rejected request goes back for re-allocation.
- Once an agency accepts, the FBO sees its details, pays, and uploads the payment proof on FoSCoS.
- The agency schedules the audit, assigns an auditor, and submits the report through AMS after its own review.
- The audit fee is standardised at ₹10,000 plus GST, excluding travel and stay.
- Audits already assigned to auditors before 19 November may be completed under the old process. No fresh manual assignments are allowed.
To be precise: this is faceless allocation, not faceless auditing. An auditor still walks the plant floor, opens the CIP records and checks the cold room. What has gone is the FBO's power to choose who that auditor works for.
Why this matters more in 2026 than it would have in 2018
Under the Licensing and Registration amendments that came into force on 1 April 2026, licences are perpetual and inspections are risk-based. How often a unit is inspected now depends on its compliance history, its test results and its audit outcomes. The Commissioner of Food Safety or the Food Authority can also direct a business to undergo a third-party audit.
That changes the weight of an audit report. It is no longer a certificate for the reception wall. It is an input that decides how often the regulator comes to the door. When the audited business chooses the auditor, and the auditor's repeat business depends on that choice, the risk score is compromised before it is calculated.
I am not suggesting that recognised agencies have been writing friendly reports. Most have not. But a regulatory system should not depend on the good conduct of the people inside it. It should be designed so that good conduct is the easiest path. Fixing the fee at ₹10,000 takes away the other lever: agencies can no longer compete for clients on price, or quietly on leniency.
What FSSAI must get right
I compliment FSSAI on this step. A faceless audit system is only as good as the rules behind it, though. Four things will decide whether it works:
- Publish the allocation logic. Businesses and agencies should know the basis on which assignments are made: rotation, geography, capacity, scope. An algorithm nobody can see is just a new kind of discretion.
- Match scope, not just names. A dairy plant with UHT lines, CIP systems and a cold chain needs an auditor who has worked in dairy. AMS already records each agency's scope of auditing. That scope must govern allocation strictly.
- Watch the rejections. If agencies can refuse assignments freely, they can return to choosing their clients by declining the ones they don't want. Rejection rates by agency should be monitored and capped.
- Audit the auditors. FSSAI now holds both third-party audit scores and its own enforcement findings for the same units. Where an agency's audit scores consistently fail to predict what inspectors and laboratories later find, that agency should be reviewed.
The remaining link is payment. The FBO still pays the agency directly. With a fixed fee and system-driven allocation, that link matters far less than before, but over time payment through FSSAI's own platform would close it completely.
From faceless audit to faceless sampling
On 12 July 2026, I wrote in these columns, in Trust, Not Transfers: Why India's Food Safety System Needs a Faceless Sampling Model, that food safety enforcement in India depends too heavily on which officer holds which post in which state. Strong officers produce raids and headlines. Their transfer often ends both. Consumers cannot build trust on that.
I proposed three changes:
- Every fourth enforcement sample should be tested at an independent NABL-accredited laboratory. Neither the officer nor the FBO should choose the laboratory.
- Centrally generated barcodes should replace business identities on samples, so the laboratory tests the sample and not the name.
- Samples should be dispatched to laboratories by the system, not through district-level arrangements, with a second round of re-coding so that even the analyst does not know whose sample it is.
FSSAI has now applied the same principle to auditing. FBOs no longer decide which agency audits them. My July proposal said that officers should not decide which laboratory tests their samples. The logic is identical: remove human discretion at the point of allocation.
The stakes in sampling are higher. An audit leads to a report. A sample can lead to prosecution. If faceless allocation is right for the lower-stakes activity, it is right for the higher-stakes one.
Faceless sampling will not be random sampling
The objection I hear most is that sampling is too scientific to be allocated by software. I agree that it cannot be allocated randomly. It must be allocated by rules.
Take milk. A sample drawn for coliform or standard plate count must reach the laboratory cold and within hours. A ghee sample for lead or a milk powder sample for melamine has no such urgency, but needs a laboratory with the right instrument and accredited scope. Any faceless system must factor in accreditation scope, distance, perishability, chain of custody and turnaround time. It must also fully preserve the statutory sampling procedure and the FBO's right to have its counterpart sample tested.
None of this is beyond FSSAI's existing digital stack. FoSCoS, FoSCoRIS and InFoLNet already carry most of the data needed. What is missing is the decision to connect them.
That is why I proposed starting with a pilot in a few states with high sample volumes, publishing the allocation rules, and reviewing outcomes after a year before any national roll-out.
Food Safety 2.0: from officers to institutions
Food Safety 2.0, as I have described it, is not about adding more portals. It is about going through every stage of food safety administration and asking one question: does this outcome depend on who happens to be in the chair?
If it does, the system should be redesigned so that it does not. Inspection assignment, sample allocation, laboratory identity and chain of custody all need to pass this test. So does the way we measure enforcement, which, as I argued in Beyond Raids: Should Regulators Also Be Measured by Prevention?, should count compliance improved rather than raids conducted.
Good regulation has two jobs. It must protect consumers from unsafe food. It must also protect compliant businesses from arbitrary enforcement. A transparent, rule-based system does both. Faceless processes do not soften enforcement. They make it harder to dispute, because nobody can claim the result depended on who chose the lab.
Today FSSAI decided that a food business should not pick its auditor. The next step is to decide that an officer should not pick the laboratory.
The FSSAI faceless audit is the first brick. Faceless sampling should be the next.
References
- FSSAI AMS: Implementation of System-Based Allocation of Auditing Agencies (effective 19 November 2026)
- Kuldeep Sharma, Trust, Not Transfers: Why India's Food Safety System Needs a Faceless Sampling Model, DairyNews7x7, 12 July 2026
- Food Safety and Standards (Licensing and Registration of Food Businesses) Amendment Regulations, 2026, Gazette notification dated 10 March 2026 (in force 1 April 2026)
Never miss an important dairy update
Follow DairyNews7x7 on Google to get live milk prices, policy alerts, and daily headlines directly in your feed.







