It is a Thursday afternoon at an invented manufacturing company in Coimbatore, and the security lead has booked a conference room for two hours. Around the table sit the head of IT, the plant operations manager, someone from legal, the HR manager, the communications officer, and a finance controller. Nobody has a laptop open. The lead reads from a sheet: “At 6:40 this morning, the shift supervisor reports that the production planning screens show a ransom message. Several file shares are inaccessible. What happens now?”

That is a tabletop exercise. It costs almost nothing, and in my view it is the most underused control in security. Follow this room for a while.

What happens when the plan meets people

The facilitator reveals the scenario in stages, called injects. Each inject adds a complication, and the group says what they would do, who decides, and who they would call.

Inject one: IT says the file server is encrypted. Who may authorise taking the network offline? The plant manager says production cannot stop without his approval, and IT says it needs to isolate immediately. Nobody knows who outranks whom in this situation. The written incident response plan, it turns out, says “escalate to management” without a name.

Inject two: a message from the attacker claims customer data has been copied and sets a deadline. Legal asks what the company’s obligations are. In India, that leads to questions about reporting to CERT-In within the six-hour window that its 2022 directions set for certain incident types, and to the DPDP Act’s expectations around personal data breaches. Legal does not have CERT-In’s contact details in the plan, and nobody is sure who files.

Inject three: a journalist emails asking about rumours. The communications officer realises there is no approved holding statement, and wonders whether employees may talk to anyone at all.

Inject four: finance asks whether the company’s cyber insurance policy needs notification before engaging an external forensic firm. Nobody has read the policy conditions.

Two hours later the whiteboard holds a list of eleven gaps. Not one involves a firewall setting. They are about authority, contacts, decisions and sequencing, which is where real incidents usually get stuck.

Running one well

  • Choose a realistic scenario from your own risk assessment: ransomware, a mailbox takeover that leads to payment fraud, a lost laptop, a supplier breach.
  • Invite decision makers, not only technicians. A tabletop attended only by IT tests only IT.
  • Make it blame-free. If people fear looking foolish, they say what the plan says rather than what they would really do.
  • Keep a scribe and finish with an action list, each item with an owner and a date. Without follow-up it was a pleasant meeting.
  • Repeat with different scenarios and after major changes, such as a new cloud platform or a new head of department.

Limits: a tabletop tests decisions, not tools. You will not find out that your backup restore takes nine hours instead of two. For that you need technical recovery tests.

From the conference room to the keyboard: purple teaming

Back in Coimbatore, the next quarter’s exercise is different. Here is the vocabulary. A red team plays attacker, emulating the behaviour of real adversaries. The blue team defends: monitoring, detection, response. In a classic red-team engagement, the blue team does not know it is happening, and the result is a report at the end. Purple teaming is the cooperative version: red and blue work together, in the same room or the same call, and treat the exercise as a way to improve detection directly.

The loop is short. Pick a technique from a framework such as MITRE ATT&CK, for example a way attackers dump credentials or create persistence. The red side performs a safe, agreed version of it in a test environment or on approved systems. The blue side checks: did our tools log it? Did an alert fire? Was it understandable? If not, they write or tune a detection, and the red side repeats the technique to confirm it now works. Then they move to the next technique. After a day, the team has a measured list: techniques covered, techniques detected, gaps closed.

The merit is speed of learning. A conventional test might tell you after six weeks that “lateral movement was not detected”. A purple session fixes that detection by lunchtime. Some organisations also use automated breach and attack simulation tools to run such tests regularly, which is useful for regression checking but does not replace human creativity.

Which one, when

If you have never rehearsed, start with a tabletop. It is cheap, quick, and uncovers organisational gaps. Once you have a functioning SOC and a set of detections, add purple teaming to harden them. A covert red team engagement makes sense when you are reasonably mature and want to test the whole chain, people and process included, under realistic pressure.

Whichever you choose, document what you learned, and close the gaps before the next one. An exercise that finds the same eleven problems twice has measured your follow-up, not your security.

Figures and names in the examples are illustrative.

Leave a Reply

Trending