Stories

Coles and the Cyber Incident: What "No Evidence of Compromise" Really Means

Published 16 November, 202614 min read

On the morning of 16 November 2026, Coles Group told the Australian Securities Exchange that it was investigating a cyber security incident. Within hours, the statement had been rewritten by headline writers as a hack, a breach, and an attack, depending on the outlet. The company's own words were more careful, and that gap between the filing and the coverage is where the useful lesson sits.

Coles said it had not identified any evidence that customer, team or supplier data had been compromised or removed from its systems. It said stores and online shopping were operating normally. It said it had engaged the Australian Cyber Security Centre and unnamed global cyber security experts to contain the incident and secure its IT environment. It did not say what kind of attack it faced, how intruders got in, which systems were touched, or when the problem was detected.

For a company of Coles' scale, that is a meaningful disclosure and an incomplete one at the same time. This article unpacks what the retailer actually confirmed, why the phrasing matters more than it looks, what the operational signals tell us, and what shoppers and suppliers should realistically do while the investigation runs.

What Coles actually disclosed

The core of the announcement is short enough to hold in one breath. Coles is investigating a cyber security incident. Its investigation to date has found no evidence of compromised or removed customer, team or supplier data. It is working with the ACSC and will keep engaging with appropriate agencies and regulators. It has brought in outside experts to contain and secure its environment. Trading continues.

Because Coles is listed on the ASX, that statement doubles as market disclosure. That changes the register of the language. Listed companies draft these filings with legal counsel in the room, so verbs are chosen rather than reached for. Coles wrote "investigating," not "contained." It wrote "not identified any evidence," not "no data was taken." And it wrote "to date," which quietly keeps the door open for the investigation to find more.

A second layer of detail appeared on the company's public updates page rather than in the exchange filing. Coles listed gift cards as working normally, said there was no indication Flybuys accounts had been affected, and said no issues had been identified with Coles Plus, Coles Plus Delivery or QuiteLike subscriptions. It also acknowledged a small number of customers experiencing problems with Liquorland click-and-collect, which it said it was investigating.

That last line matters more than its placement suggests. Liquorland, Coles' liquor business, reportedly suspended click-and-collect temporarily while it worked through the incident, telling customers in an email that stores remained open and home delivery was continuing. Neither Coles nor Liquorland specified the cause or a restoration timeline.

The confirmed and the unknown

Early incident coverage routinely blurs two categories: what the company said, and what independent parties have verified. Separating them is the fastest way to size an incident honestly.

Item Status on 16 November 2026
Coles is investigating a cyber security incident Confirmed by Coles
No evidence customer, team or supplier data compromised or removed Coles' finding to date
ACSC involvement Confirmed by Coles
Global cyber security experts engaged Confirmed by Coles
Stores and online shopping operating normally Confirmed by Coles
Type of attack Not disclosed
Attack vector and affected systems Not disclosed
Threat actor or extortion demand Not disclosed
Whether any data was accessed Not established
Detection timing Not disclosed

Five rows sit in the confirmed column and five in the open one. That split is ordinary for a first-day disclosure, and it explains why several outlets described the situation as one with few details available. The facts that would let a reader judge severity are precisely the ones being withheld, most likely because they are not yet known internally.

Why "no evidence" is not the same as "no breach"

Security teams treat "we have not identified evidence" as a statement about an investigation, not about the world. It describes what responders have found so far. It does not describe what an attacker did.

The distinction is not pedantry. Across industries, early "no evidence" language has often given way to a later "some data was accessed" update once forensic work reached deeper logs, archived backups, or third-party systems that had not yet been examined. An investigation that is three days old and one that is three weeks old can produce genuinely different answers from the same evidence base.

Coles' wording is consistent with several scenarios at once. An attacker may have been detected and stopped before reaching data stores. A system may have been disrupted without any copying of files. Or the response team may simply not have finished checking. The public statement gives no way to tell these apart, and no responsible reader should claim either that Coles' data is safe or that it was stolen.

The word "compromised" also deserves a second look. Data can be viewed, altered, or encrypted without being removed, and the filing pairs the two verbs with "or." As written, the sentence covers both, which is broader than the headline summaries suggested. What it does not cover is the full estate. Customers, team members, and suppliers are the people whose records a large retailer holds, but the statement says nothing about commercial contracts, internal financials, or operational technology data. That is not evidence of anything. It is a reminder that the sentence is scoped to people's personal and business information, not to every file on every server.

There is also a timing signal worth noting. The disclosure was proactive. Coles went to the market before anyone forced it to, and it named the incident as an incident. Companies that are trying to bury a problem usually do not volunteer a category as loaded as "cyber security incident" on a Monday morning.

Operations continuing is a real signal, but a limited one

Coles said customers could shop as normal in stores and online, and that no action was required of them at that time. AAP's coverage paired the reassurance with a warning to be vigilant, which is a fair summary of the company's position: keep shopping, keep watching your inbox.

For an incident responder, continued trading is informative. Ransomware crews often aim squarely at availability, because downtime creates the pressure that produces payment. If supermarket tills, the Coles app, the website, and online fulfilment are all still running, that points away from a disruptive, encryption-first attack. It does not rule one out. Coles has said nothing about extortion demands, and the filing is silent on whether ransomware is involved or excluded.

The Liquorland click-and-collect disruption complicates the picture slightly, and it is the one concrete operational symptom in the public record. A single degraded service inside a large group is consistent with containment measures, with a compromised integration, or with an unrelated fault that happens to be under investigation at the same time. Coles has not said which, and readers should resist the temptation to fill that gap with the most dramatic available explanation.

Who is working the response

Two outside parties appear in the disclosure. The first is the Australian Cyber Security Centre, the government body that coordinates cyber incident reporting and response. Coles said it is working with the ACSC and will continue engaging with appropriate agencies and regulators, which is standard for an ASX-listed company handling a notifiable event.

The second is a group of unnamed "global cyber security experts" engaged to contain and secure the Coles IT environment. Large retailers typically hold retainers with digital forensics and incident response providers, so a rapid call-up is routine practice rather than a marker of unusual severity. The absence of names is normal at this stage; forensic providers frequently prefer not to be publicly attached to an active investigation.

The word "contain" is worth pausing on. Containment means limiting an intruder's reach. Securing means hardening what remains. Together they describe an active response rather than a closed file, and they sit comfortably alongside the "investigation to date" framing. The language implies Coles expects to learn more in the coming days.

The wider context: a bad season for Australian organisations

The Coles disclosure did not arrive in a quiet month. Australia has spent much of 2026 working through a dense sequence of incidents across retail, health, government, aviation, and professional services.

In July, Origin Energy disclosed that hackers had taken data belonging to around 900,000 current and former customers, including concession card details for up to 15,000 people and full bank account numbers for a much smaller group. In November, Qantas was notifying customers about a breach at third-party travel technology provider Amadeus. NSW Health dealt with unauthorised access to an isolated server tied to a third-party dental provider scheme. The pattern across all of them is the same one Coles is now navigating: the first statement tells you what still works, and the difficult answers arrive later.

The closest analogy for a supermarket, though, is not a data breach at all. In October, hackers manipulated refrigeration systems across hundreds of European grocery stores, driving temperatures up and spoiling food worth tens of millions of euros. France's national cybersecurity agency linked the incident to outdated firmware in refrigeration controllers supplied by Danfoss. Later reporting indicated that similar controller families and remote management platforms are used in Australia, which prompted Woolworths and Aldi to state that their refrigeration controllers are not exposed to external networks. Coles declined to answer questions on the subject at the time.

That episode matters here because it reframes what a supermarket cyber incident can be. The traditional mental model is stolen personal information. The emerging one is operational technology: the systems that keep food cold, warehouses moving, and shelves stocked. As one security researcher put it, attackers increasingly are not just going after data, they are trying to shut things down. Coles has described an IT environment issue, not a refrigeration issue, and nothing in the public record connects the two. But the European case explains why retail incident disclosure now attracts attention well beyond the privacy community.

What customers should actually do

Coles' advice to customers was deliberately modest: keep shopping, stay alert to suspicious messages, and use a strong, unique password for your account. That is generic guidance, and it is also the highest-value action available to an individual after any retailer incident.

Password reuse is the mechanism that turns one company's bad day into a problem across every service a person uses. Attackers know this and routinely take credential lists from one breach and test them against unrelated sites. A password manager removes the friction from unique passwords, and enabling two-factor authentication wherever a retailer account offers it adds a second barrier that a leaked password cannot cross.

The more immediate risk, though, is phishing rather than credential stuffing. High-profile incidents create perfect cover for scams, because people are primed to expect communication from the brand. Fraudsters exploit that expectation with fake security notices, fake refund offers, and fake delivery problems. Coles stated plainly that it will never ask for a password or sensitive account information through an unsolicited email or SMS. Anyone who receives a message that does should treat it as hostile, regardless of how convincing the branding looks.

The practical habit is simple: never log in through a link in an unexpected message. Open the app or type the address yourself. If a message claims there is a problem with an order or an account, check the account directly. If it still looks wrong, call the customer response line rather than replying to the message.

Action Why it helps Effort
Set a strong, unique Coles account password Limits credential reuse and matches Coles' own advice Minutes
Replace reused passwords on other sites Stops one incident spreading across accounts An afternoon
Ignore links in unexpected Coles emails and texts Defeats the phishing wave that follows publicity Ongoing habit
Open the app or site directly to check messages Bypasses counterfeit login pages Seconds
Watch bank and card statements Catches misuse early if payment data is ever involved Minutes a week

The final row is general hygiene rather than a response to this incident. Coles has not suggested payment data is involved, and the disclosure points the other way. Reviewing statements costs almost nothing, though, and it is the only way an individual would notice card misuse early.

What suppliers should watch for

Most breach statements mention customers and employees. Coles named suppliers as well, and that choice says something about where the company sees risk. A national grocer sits at the centre of a web of producers, distributors, logistics partners, and service vendors. Contracts, invoices, remittance details, and bank account information flow through the retailer's systems. A leak there can damage small businesses that have no security function of their own.

Supply chain exposure has already shaped several incidents this year, most visibly the Amadeus breach that pulled airlines including Qantas into notification duties, and the Retail Express incident that exposed roughly 339,000 customer records held by cosmetics brand Napoleon Perdis. In both cases the organisation facing the public was not the one whose systems were first compromised.

For suppliers, the sharpest near-term risk is not data theft but business email compromise. Fraudsters routinely exploit news of an incident to send fake payment change requests, on the assumption that finance teams are already expecting disruption and may be less careful than usual. Any sudden request to change bank details connected to a Coles account should be verified through a known phone number, never through the contact details in the email that made the request.

How this story is likely to develop

Three things will determine whether the Coles incident becomes a footnote or a significant event.

The first is the forensic timeline. If responders can show that the intrusion was detected early and contained before reaching data stores, the "no evidence" framing will hold and the story will fade. If deeper log review surfaces access to customer or supplier records, Coles will have to update the market, and the phrase "to date" will do a lot of quiet work in that follow-up.

The second is attribution and motive. No threat actor had claimed responsibility at the time of the disclosure, and Coles has not described an extortion demand. A financially motivated intrusion, an espionage attempt, and an unattributed probe all produce different disclosure obligations and different public interest.

The third is whether the operational disruption spreads. Liquorland click-and-collect is currently the only acknowledged service impact. If additional services degrade over the following days, the reassuring "shop as normal" message will need revisiting.

Reading the next update properly

Incident disclosures are not binary. They are a sequence, and each statement is a snapshot of what the company knew at the moment it published. The right way to read Coles' first statement is as a narrow factual claim, honestly made, about an investigation that had barely begun.

That framing cuts both ways. It is not grounds for panic, because there is no evidence in the public record that anyone's data has been taken. It is also not a clean bill of health, because the investigation that would produce such a conclusion is still running. What Coles has told the market is that its stores are open, its response is staffed, regulators are informed, and its forensics have not yet found data loss. Every one of those statements is verifiable and every one of them could change.

For shoppers, the practical answer is unchanged: keep using the app and the website, keep an eye on your accounts, and treat unexpected messages with suspicion. For suppliers, it is to confirm payment details out of band and to assume that fraudsters are already drafting emails that reference this incident. For everyone else watching how Australian companies handle cyber disclosure, Coles has done the straightforward part well. The hard part, the part that determines what this incident actually was, is still being written.

References