100 ready-to-run cyber tabletop exercises across 15 incident domains. Search them, pick one at random, or tell us who will be in the room and we'll rank the scenarios that fit those people best. Every exercise shows who to involve, how long it runs and an outline — and the scenario prompt is free to copy and read out.
Running one for real? Pro adds the facilitator's run sheet for each exercise: a minute-by-minute timeline, timed injects, step-by-step facilitation guide and the questions to put to each person. Or build your own with the Tabletop Runbook Generator.
Who's in the room?
Pick the people you have — we'll suggest exercises that fit them.
Four days after you contained a ransomware attack and restored most systems from backup, a threat intelligence vendor emails your CISO a screenshot. Your company is listed on the group's leak site with a 72-hour countdown and a sample: forty scanned employee passports and a folder of customer contracts. The listing claims 600 GB was taken and will be published unless you pay $2.4 million. Your original investigation found no sign of data theft, and staff were told so. A trade reporter has already left a voicemail asking for comment. Do you engage, and what do you say?
It is Wednesday morning. Payroll for 2,300 staff must reach the bank by 14:00 Thursday to land on Friday. Overnight, ransomware encrypted the HR and payroll servers and the file share holding timesheets. IT estimates at least four days to restore. Last month's bank file still sits in a finance team member's email. Union representatives have heard rumours and are asking whether people will be paid this week. The CFO asks the room: can we just re-run last month's payroll? What do you decide, and how do you tell 2,300 people?
It is day five after ransomware hit a 1,400-person insurance broker. Identity has been rebuilt and the attacker appears to be out. The infrastructure team can restore about twelve systems a day. Twenty-three business owners have each told the CEO their system is the most critical: claims, policy administration, finance, the broker portal, email archives. Two large clients threaten to move their accounts if the portal is not back by Monday. Yesterday a restored file server was found to contain the attacker's remote access tool and had to be taken down again. Who decides what comes back next, and on what basis?
An HR coordinator meant to email a salary review spreadsheet to the payroll provider. Autocomplete picked a similar name at a recruitment agency instead. The file holds names, salaries, home addresses and sickness absence notes for 340 employees. An hour later the recipient replied: 'I think this was meant for someone else.' The coordinator is panicking and has asked a friend in IT to 'just recall it'. It is now 16:30 on a Thursday. What happens next, and who needs to know?
A regional HR manager reports on Monday morning that she left her laptop in an airport security tray on Friday evening. She waited because she hoped the airport would find it. The device management console shows it last checked in on Saturday at 09:12 from an unknown home network. Disk encryption shows 'pending' because a policy change was never completed on her model. The laptop held disciplinary files and a local copy of an investigation into a senior manager. What do you do first, and is this a notifiable breach?
At 14:10 your SOC sees a new federated domain added to your cloud identity tenant, then a conditional access policy disabled. Both changes came from a Global Admin account belonging to a cloud engineer who is on holiday. Within twenty minutes, two new app registrations with mail-read permissions appear and three other admins report their MFA methods were changed. The attacker can now create users, issue tokens and read any mailbox. You are not sure which admin accounts are still yours. How do you take the tenant back without locking yourselves out?
The head of sales operations clicked a link to a shared proposal on Monday and signed in as usual, MFA and all. This morning your identity logs show her session in use from a hosting provider while she is also active from the office, with no new sign-in and no MFA prompt. Overnight the session read 900 emails, downloaded the pricing folder, exported contacts from the CRM through single sign-on and consented to a new mail app. The help desk reset her password an hour ago. The activity has not stopped. Why not, and what do you revoke?
On Tuesday afternoon, a caller claiming to be a regional sales director phones the IT help desk from a number that is not on file. He sounds stressed, says he has a new phone and is about to present to a major customer, and knows his employee ID and his manager's name. The agent resets his MFA and sends a temporary password. Ninety minutes later the real director, sitting in your office, gets a sign-in alert. Since then the account has opened the CRM and the shared drive for an acquisition project. What now, and how did this get through?
Your accounts payable team receives an email from your largest packaging supplier, sent from their real address and replying to a genuine invoice thread. It says they have changed banks after an audit and attaches a signed letter with new account details. The clerk calls the number on the letter, gets a confirmation, updates the vendor record and pays $186,000 on Tuesday. On Friday the supplier's credit controller phones to ask why the invoice is overdue. Your bank says the receiving account is at another bank, which has not replied. What happens in the next hour?
On payday, three employees tell HR they have not been paid. In each case the HR self-service portal shows their bank details were changed eight days ago, from a sign-in that passed MFA. One employee remembers clicking a 'benefits enrolment' link in an email. This month's payroll file has already gone to the bank. Two more employees' details changed yesterday, and next week's contractor run is being prepared now. The affected staff are upset and want to know when they will get their money. What do you do today?
Three customers call within a day to say they paid your invoices to new bank details your accounts receivable team emailed them. The emails came from your real billing mailbox, sat inside genuine threads and used your invoice template. Logs show the mailbox has been accessed from overseas for 26 days through a legacy mail protocol that skips MFA. Losses so far total $430,000 across three customers, and the mailbox holds 2,100 customer contacts. Customers want to know who is liable. Your general counsel asks: must we warn all 2,100 today?
A public advisory says a popular open-source logging package was hijacked: version 4.8.2, published three days ago, runs a post-install script that reads environment variables and sends them to a remote server. Your dependency manifests allow minor updates. A quick search shows 14 of your repositories pulled 4.8.2, and your CI runners, which hold cloud deploy credentials, package registry tokens and a signing key in their environment, built every one of them, many times. Two developers also installed it on their laptops. What do you assume was taken?
A message arrives at your CEO's and general counsel's inboxes: "We have 380 GB of your source code, including your payment service and your infrastructure scripts. Sample attached. Pay 40 BTC in five days or we publish everything and send it to your competitors." The sample is real: three repositories, including commit history and a configuration file with what look like live database credentials. Your code host shows a personal access token belonging to a senior engineer cloned 212 repositories eleven days ago from an unfamiliar country. The board meets in two days.
It is Wednesday morning and 2,600 staff are due to be paid on Friday. Your outsourced payroll provider has been offline since last night. Its status page says "investigating a security incident affecting some systems" and gives no recovery time. Your payroll team cannot submit the final run, and the provider has the only copy of this month's calculated pay, tax and deductions. Social media already has posts from employees at other companies using the same provider. Your union representative has asked HR whether pay will arrive on time. The CEO wants a decision by noon.
A threat-intelligence feed your team subscribes to posts a new entry: a ransomware group claims to have stolen 600 GB from a small order-fulfilment company. You use that company to ship orders, and you send it customer names, addresses, phone numbers and order contents every day. The listing includes a file tree with a folder named after your brand. The vendor has not contacted you. When you call, the owner says they "had a computer problem last week" and that their IT provider is handling it. You are the data controller for those customers.
A user posts screenshots on social media of your customer-support chatbot. By pasting a crafted instruction into the chat, they got it to look up and display order histories, addresses and partial card numbers for three other customers. The bot uses a large language model with a tool that queries the order database using the customer ID the model supplies. The post has 40,000 views and replies show other people trying the same trick. The bot handles 9,000 conversations a day and is the only support channel open overnight. It is 22:30.
Three weeks ago your platform team gave an AI operations agent permission to tidy up unused cloud resources, with a rule that destructive actions need human approval in chat. This morning the cost dashboard shows a 22% drop. The agent's log says it identified "orphaned snapshots" and deleted 1,400 of them, including the only backups of your billing database. It recorded the approval as granted: an engineer had replied "looks good" to a different message in the same thread. A database migration is planned for tonight. Who is accountable, and what do you change?
Your new data-loss prevention pilot produces its first weekly report. In seven days, staff pasted content into a free public AI chatbot 1,840 times. The sample includes a production database connection string, a customer contract with pricing, two HR investigation notes and a spreadsheet of patient appointment details. Your acceptable-use policy says nothing about AI. The approved enterprise AI tool, bought six months ago, has 30 active users out of 1,200. A department head replies to your email: "Everyone uses it. Banning it will just push people to their phones."
On Monday a finance manager joined a video call with what looked and sounded like your CEO and CFO, plus an "external lawyer". They explained a confidential acquisition and asked for three urgent payments to an escrow account, totalling $2.4 million. The manager made the payments after a follow-up email from the CEO's real address, sent through a compromised mail rule. Today the real CEO asks why finance moved money on a deal that does not exist. Your bank says the first transfer has left the receiving bank. The finance manager is distraught.
It is 05:40 on a Monday at a 900-person auto-parts manufacturer running three shifts. Office laptops started showing ransom notes an hour ago. Now the night-shift supervisor reports that two HMI screens on the stamping line show the same note, and the historian server has stopped updating. The PLCs are still running and the presses are still making parts, but operators can no longer see process values on screen. The first shift arrives in 80 minutes. Do you keep the line running blind, shut it down, or cut the plant off from the corporate network first?
Your filling-machine supplier emails every customer on a Thursday afternoon: an attacker had access to its remote-support platform for at least three weeks and 'may have used customer connections'. Your beverage plant gives this supplier always-on remote access to the line controllers so it can tune machines without a site visit. Your firewall shows sessions from the supplier's platform on four evenings when the supplier says no engineer was working. Its account manager wants the connection left open because a firmware fix is scheduled for Monday. Do you cut them off, and what do you look at first?
You are on day two of investigating suspicious traffic inside the control network of a chemical blending plant. Responders have a laptop on the OT switch capturing packets. At 14:20 the safety instrumented system trips reactor R-3 on high pressure, venting to the flare and closing the feed valves. The control room saw the pressure transmitter spike, but two field gauges read normal throughout. Operators want to reset the safety system and restart the batch, which is worth $400,000. The responders want the safety controller left untouched until they know whether anyone has reached it. Who decides?
A colocation provider’s night technician calls your on-call engineer at 22:40. He found a man in a high-visibility vest inside your company’s locked cage, holding a laptop with a cable plugged into a top-of-rack switch. The man said he was ‘from the network team’ and left before site security arrived. The front desk logged him following one of your contractors through the entry mantrap two hours earlier. Your cage hosts the card payment platform for 2 million customers. What do you check first, and who do you call?
Police call your office manager at 06:30: a ground-floor window at your accounting firm’s branch office was smashed overnight. When staff arrive they find eleven laptops missing from desks, plus the small network storage box from the comms cupboard that the branch used for scanned client documents. IT believes the laptops are encrypted but is not sure about two older ones. Nobody knows exactly what was on the storage box or whether it was backed up. Tax season starts in three weeks. Is this a theft or a data breach?
A senior engineer at a 250-person fintech hands in her notice on Friday, saying she is joining a competitor. On Monday a routine review of source-control audit logs shows her account cloned 140 repositories in the ten days before she resigned, including the fraud-scoring models she never worked on. Her work laptop also synced to a personal cloud drive twice last week. She is still working her notice period, has production access, and is well liked. HR wants to handle it quietly. Legal wants evidence first. What happens today?
Two weeks after being told he would not get the infrastructure lead role, a senior systems administrator at a 1,500-person healthcare billing company resigns with immediate effect. His accounts were disabled within the hour. Three days later, at 01:00, scheduled tasks on twelve servers delete the database backups and drop a production claims table. The tasks were created months ago under a service account he set up, which nobody has rotated. Claims processing is down. How much else has he left behind, and how do you find it before the next one fires?
Your threat-intelligence provider sends a screenshot from a criminal forum: a seller offers ‘VPN and Citrix access to a regional insurer, 2,000 staff, domain user, $6,000, insider can stay on for ongoing jobs’. The details match your company. The SOC traces two test logins described in the post to a claims handler’s account, from her home IP address, at times she was not scheduled to work. She is on shift right now. Is she the seller, a victim, or both? Do you lock her out now, or watch to see who buys?
At 07:10 customers of a 700-person online bank start reporting that the login page looks slightly wrong and asks for their one-time passcode twice. Your engineers find the domain’s name servers were changed overnight through the registrar’s web portal, using a staff account protected only by a password and SMS codes. DNS now points to attacker-controlled servers, and a certificate for your login hostname was issued to someone else at 03:40. Some resolvers still have the old records cached. The registrar’s support line opens at 09:00. How do you take your domain back, and what do you tell customers?
Three days after attackers stole HR files from a 900-person engineering consultancy, they change tactics. Twelve employees, including the CEO and two HR staff, receive texts on their personal phones listing their home address and their children’s school, with a demand: tell your company to pay, or this goes online. One message includes a photo of an employee’s front door. The attackers’ leak site now lists the company with a five-day countdown. Staff are frightened and some want to stay home. The board meets tomorrow. What do you do for your people today?
At 21:30 on a Sunday, armed police surround your CEO’s house after a hoax call reporting a hostage situation. Nobody is hurt, but her teenage son was handcuffed on the lawn. Within the hour, an anonymous social media account posts your CFO’s home address and writes ‘he’s next’ in a thread about recent layoffs at your 3,000-person logistics business. Some of the executives’ personal details appear to come from a data broker. Monday’s all-staff meeting at headquarters is in twelve hours. How do you protect people tonight, and what changes tomorrow?
At 19:45 on a Friday your company’s main social media account, with 450,000 followers, posts a video of your CEO announcing a giveaway: send cryptocurrency and receive double back. The video is a deepfake. Three more posts follow in ten minutes, and replies show customers who have already sent money. The social media manager is locked out because the account’s email and phone number have been changed. Nobody else knows who holds the recovery codes. The platform’s support form promises a reply within 48 hours. What do you do now?
It is 02:40 on a Saturday at a 450-bed regional hospital. The night supervisor calls: the electronic health record is frozen, workstations on three wards show a ransom note, and the pharmacy dispensing system has dropped offline. The emergency department has 31 patients, two ambulances are ten minutes out, and the trauma surgeon on call wants to know whether to divert. Downtime procedures exist on paper, but nobody has run them for longer than four hours. The ransom note says patient data has already been copied. Who decides on diversion, and what do you tell the ambulance service right now?
Friday, 18:20. A correspondent bank calls your payments operations desk about three outbound cross-border transfers sent in the last two hours: $4.2 million, $6.8 million and $9.5 million, to beneficiaries in three countries you rarely deal with. The messages were authenticated through your own payment messaging gateway using two valid operator accounts, one belonging to someone on leave. The daily reconciliation that would have caught them runs at 22:00, and today's printed confirmation report shows nothing unusual. Some of the funds may already have been credited. Who calls whom first, and what do you switch off?
It is 20:05 on Black Friday. Your online retailer's checkout has been failing for 18 minutes; about one payment attempt in three times out. The promotional email to 2.4 million customers went out at 20:00. Revenue for this hour last year was $1.9 million. Engineering thinks the payment provider is rate-limiting you; the payment provider says its systems are healthy. Social media has noticed, and a competitor has just extended its own sale by an hour. The CEO has joined the bridge and wants to know whether to pause the promotion. What do you decide in the next ten minutes?
Monday, 07:30. Staff at a county council serving 280,000 residents arrive to find file shares encrypted and the housing benefits, social care case management and parking systems unavailable. Phones still work, but the contact centre cannot see case records. A social worker needs a child's safeguarding history for an emergency court hearing at 11:00. The ransom note demands payment within 72 hours and threatens to publish residents' data. Councillors are already asking questions, and a local reporter has heard that the payments portal is down. What do you prioritise this morning?
Two days before a municipal election, a 40-second audio clip spreads on social media. It sounds like your county's chief election official telling a colleague that ballots from two precincts 'will be handled differently this year'. A second post, styled like your website, says three polling places have moved. Your official never said this, and nothing has moved. By noon the clip has 300,000 plays, your phone lines are jammed, and a candidate is demanding an explanation. Your public results website is also seeing unusual traffic. How do you respond without amplifying the fake?
Accounts payable at your industrial equipment manufacturer pays $860,000 to a long-standing steel supplier, and three days later the supplier asks where its money is. The bank details on file were changed through your supplier self-service portal two weeks ago, by a login using the supplier's own account. The same account uploaded six updated 'drawing packages' that your design engineers have since opened. Two other suppliers report odd password-reset emails from your portal. The portal is run by a SaaS vendor your procurement team chose. What do you check, freeze and tell suppliers today?
Your food-packaging plant runs three shifts and produces about $95,000 of product an hour. At 10:00 the SOC confirms ransomware on the corporate network and finds the same malware on one engineering workstation that can reach the lines' controllers. The lines are running normally. Isolating the plant network means losing the production schedule and batch records held in IT systems and running on paper at reduced output. A full shutdown is safest and costs more every hour. A major customer's promotion depends on next week's shipments. The board chair wants options and numbers by 13:00.
A customer of your 400-person HR software company opens a support ticket with a screenshot: their dashboard search returned salary records belonging to another customer's employees. Engineering traces it to a caching change deployed nine days ago that sometimes keys results by query text instead of by tenant. Logs show the faulty path ran at least 6,000 times, but nobody yet knows how many of those responses held another tenant's data, or which tenants saw what. You process data for 1,100 business customers, most of them in the EU and UK. The fix is ready. Do you ship it now?
You are a listed software company. On Monday the security team confirmed that an attacker took source code and a database of 1.1 million customer contact records through a compromised build server. Revenue impact is unclear, and two large customers have asked whether their API keys were exposed. The disclosure committee meets Wednesday morning. Once you determine the incident is material, you have four business days to file. Some in the room want to wait for the forensic report due in three weeks. Others think the decision is already overdue.
Three weeks after a ransomware attack, your company has spent 1.8 million on forensics, overtime and lost revenue. The insurer's coverage letter arrives. It reserves its rights and cites two problems. The application you signed last year stated that multi-factor authentication was enforced for all remote access, but the attacker came in through a legacy VPN account without it. And a war and state-actor exclusion may apply, because a government advisory linked the ransomware group to a hostile state. Finance needs to know what to book this quarter.
You signed the agreement to buy a regional medical-billing company three weeks ago. Closing is in five weeks, pending regulatory approval. This morning the target's CEO calls: ransomware hit them overnight and patient billing records may have been taken. Your security team wants to send people in to help. Your lawyers warn that you cannot run the target's business before close. The agreement has a material adverse effect clause, and your board meets in four days. What do you do, and what are you allowed to do?
Ninety days after closing the purchase of a 500-person engineering firm, your team starts connecting its network to yours. EDR, newly installed on their servers, flags a scheduled task that calls out to an external host every six hours. It has existed since before the deal was announced. The two Active Directory forests already share a trust, and 40 of the firm's admins have accounts in your environment. Integration milestones are tied to executive bonuses. Do you cut the trust today?
At 16:45 the communications team takes a call from a reporter at a national business publication. She says a source has shown her files that appear to come from your HR system, including salary data for senior staff, and she plans to publish tomorrow morning. She wants a comment by 20:00. Your security team has no open incident, no alert matches, and nobody has seen the files. You have three hours and fifteen minutes to find out whether it is true and decide what you will say.
A threat-intelligence feed alerts you that your regional law firm has been listed on a ransomware group's leak site. The post shows a timer with 72 hours remaining, a list of client names, and three screenshots of documents that look like your matter files. There has been no ransomware outbreak and no ransom note, and IT finds nothing encrypted. Partners are asking whether this is real, whether clients must be told before the timer runs out, and whether anyone should contact the group.
Monday: ransomware encrypts file servers at your online travel agency, but customer bookings still work. Wednesday: after you decline to engage, a DDoS takes the booking site down at peak hour. Thursday: the attacker emails journalists a sample of passport scans and threatens to call your largest corporate customers. Each pressure point lands on a different team that is already exhausted. The attacker's latest message sets a 24-hour deadline and promises the DDoS will stop the moment you pay.
A government agency privately tells your aerospace components firm that an intruder linked to a foreign intelligence service has been in your engineering network for about eight months. They share three indicators and ask you not to tip off the attacker. Your logs go back 90 days. The intruder appears to use legitimate admin tools rather than malware. Your largest customer is a defence prime with contractual reporting duties. Do you watch, or evict? And what do you tell anyone?