I design and facilitate many tabletop exercises at work. The scenario I keep coming back to for community banks and credit unions is simple: your core lending vendor goes dark, and it stays dark for a week.
On October 30, Abrigo disabled its Sageworks application at approximately 11:25 a.m. ET after it “identified irregular activity in our environment.” The weekend outage hit Online Portal Now, Sageworks Analyst, and the Sageworks API. By November 2, most functions were back, except borrower-facing features (BleepingComputer). Abrigo says it serves more than 2,400 U.S. entities, a company figure for all of its products, not Sageworks alone (American Banker). It later told American Banker this was a contained data security incident involving unauthorized access. It did not publicly say whether any Sageworks client data was taken, and it has not published a root-cause write-up. Since then, I’ve heard from institutions that received notices on Friday, November 13, confirming that their client data was affected. You often have to brief the board before those details are public.
The same week, Chainalysis disclosed a security incident involving an external actor using “offensive agentic techniques,” later took products offline, and by November 4 said names, usernames, password hashes, and live tokens for all Chainalysis users had been exposed.
As I wrote in The boring list still wins, a vendor-managed system going quiet already makes a good tabletop inject. Losing the lending platform for a week is the next one.
Perhaps a better framing: vendor management checklists and SOC reports still matter, as I wrote in 2017. So does a written plan for the week the vendor can’t serve you… and a tabletop that forces the room to use it. The FFIEC Business Continuity Management booklet says a critical third-party provider can be a single point of failure if you have not considered alternative providers or other contingency plans.
Here is a 90-minute tabletop you can run as-is, called “Loan origination is down for a week.” Start Friday at 11:25 a.m. ET. The borrower portal, analyst tools, and API are down, with no confirmed root cause.
- T+0 (Friday 11:30 a.m. ET): Analyst, API, and Online Portal Now are disabled. Forty commercial applications are in flight. Who owns the call to pause or continue on paper?
- T+4 hours: A borrower emails, “I submitted docs in the portal this morning. Did you get them?” What do you tell them, and who sends it?
- T+1 day (Saturday): No restore ETA today. The board chair asks the CEO if this is a breach. What do you know, and what do you not know yet?
- T+3 days (Monday morning): Lender-facing functions are mostly back, but the borrower portal is still down. How do you take applications for the next week?
- T+3 days, afternoon: An examiner asks for your contingency plan, the last TSP-outage tabletop, and the contract clauses on RTO, notification, and data return. Can you produce them in the room?
If you want a harder inject, have security recommend rotating API keys, as researchers advised during the Abrigo event (BleepingComputer). Who approves it, what breaks, and how fast can you do it?
As I wrote in my review of Cybersecurity Tabletop Exercises, write down the gaps after the exercise, assign owners, and update the vendor playbook.
Do this Monday:
- Pick your top three lending, core, or payments vendors and write a one-page “vendor dark for 7 days” playbook for each.
- Pull the contract language on incident notification, RTO or service metrics, data return, and termination assistance, and note the gaps.
- Schedule a 90-minute tabletop with lending, ops, IT, compliance, and a business executive, using the injects above.
Run the exercise once. Fix what breaks on paper before it breaks on a Friday before a holiday weekend.