Co-op: an account reset on the help desk
In April 2025 attackers impersonated a colleague, answered the security questions and got an account reset. Defenders reacted within minutes and kept stores open, but it still took about two weeks to restore services and around 40 key pieces of equipment were rebuilt.
- Entry point: the attackers impersonated a colleague and answered security questions to get an account reset. About an hour later the account was used maliciously.
- Response: the Co-op told MPs its systems alerted teams within minutes, and defences restricted the account. Remote access and VPN were restricted on the Sunday.
- Spread: by 1 May the threat had spread to main controllers, and about 40 key pieces of equipment were rebuilt.
- Outcome: member data was accessed and services took about two weeks to restore. The Co-op says it did not pay a ransom and that its response prevented ransomware being deployed.
Timeline
Went wrong Worked Context
The help desk resets an account. The attackers impersonate a colleague and answer security questions to get the account reset. About an hour later the account is used maliciously.
Alerts fire within minutes. Staff log in within an hour of the malicious activity. The Co-op describes the next 12 to 24 hours as a firefight, with defences restricting what the account could do.
The response is stood up. Incident management starts, and remote access and VPN are restricted, which slows some colleagues' work. The National Crime Agency and the NCSC are told over the weekend.
Forensics begin. Regulators and some government departments are updated.
The threat reaches main controllers. Software in the network tries to contact a threat actor's website. Defences block it and the affected zone is paused. About 40 key pieces of equipment are rebuilt.
Member data is confirmed accessed. Names, addresses, contact details and dates of birth. Members are told through a market announcement, the website and email.
Services are restored. Distribution and food ordering had run on manual, paper-based processes for about two weeks.
What went wrong
- The help desk accepted answers as proof of identity. The attackers impersonated a colleague and answered security questions to get an account reset. The reset gave them a working account.
- The intruder got further than the first account. By 1 May the threat had spread to main controllers and equipment had to be rebuilt.
- Member data was accessed. Names, addresses, contact details and dates of birth.
- Paper fallbacks are slow. One Co-op witness called them unsustainable and said an air-gapped, segregated alternative is needed for critical processes.
What worked
- Detection. Teams were alerted within minutes and were logged in within an hour of the malicious activity.
- Segregation. The Co-op says customer-facing and online services, payments and stores stayed open because systems were segregated.
- Willingness to cut access. Remote access and VPN were restricted on the Sunday, and a zone was paused on 1 May until the root cause was found.
- Early notification. The NCA and NCSC were told over the first weekend, and regulators were updated on the Monday.
- No ransom. The Co-op told MPs it did not pay and never discussed paying, and that its response prevented ransomware being deployed.
What would have helped
This section is MGB's own analysis, not part of the source material.
- Harden help desk identity checks. Knowledge-based questions are not proof of identity. Require a call-back to a registered number or manager approval, and tie privileged and multifactor resets to stronger verification.
- Protect privileged accounts from resets. Resets of administrator and high-impact accounts should need a second person and generate an alert.
- Segment so the business can keep trading. The Co-op credits segregation for stores and payments staying open. Know in advance which systems you could cut off without stopping customers.
- Be able to restrict remote access in minutes. Practise cutting VPN and remote access, and know what breaks when you do.
- Test continuity, not just recovery. Rehearse decision-making under pressure and have a plan for staff, who also bear the strain.