Skip to content
CTS Field Notes

CTS Field Notes

Overlooked problems, assumptions worth questioning, and things worth knowing.

Practical technology judgment for business leaders. Fresh perspectives for curious minds.

Assorted Links — November 20, 2026

Permalink

Please Stop Sending Wire Transfers Without Authenticating the Request.

A wire request can arrive with a voice you recognize and someone else who seems to confirm the details. The useful question for the person releasing the money is where that confirmation came from.

In a November 5 indictment announcement, federal prosecutors alleged that defendants and their associates used AI-generated voices or images to impersonate executives. Employees were directed to people they were told were outside lawyers working on a confidential transaction. Prosecutors allege that the supposed lawyers were also impostors and that employees were directed to wire money to accounts controlled by the scheme. The charges are allegations, and the defendants are presumed innocent unless proven guilty. U.S. Attorney’s Office announcement

For an unexpected executive wire request, pause before releasing the payment. Call the executive using the number already in your company directory or contacts. Ask them to confirm the amount and recipient, then verify the receiving account through your usual payment process. If you cannot reach them through a known route, leave the wire pending. A number supplied during the original request puts you back in the same conversation, even if a different voice answers.

Permalink

The Familiar Face on the Call Was a Recording

Seeing a colleague on a video call usually settles one question: who else is there? In the campaign the FBI calls FaceStealer, that was precisely the question the people running the call wanted to settle for you. The familiar face could be a recording of a person who had been targeted earlier. The person on screen had no need to be present for the next request to sound credible.

On November 12, the Justice Department announced it had disrupted cloud infrastructure used by North Korean hackers in the campaign. The FBI attributes more than $400 million in cryptocurrency thefts over 18 months to FaceStealer. That is the reported total across the scheme. The public account describes a pattern investigators saw across cases, with different consequences for different victims. Justice Department announcement

The meeting made the request seem ordinary

The invitation could appear to come from someone the recipient already knew. Federal investigators say the attackers used accounts taken from earlier victims to approach their contacts. In the apparent meeting, a familiar face could reinforce the impression that the invitation was legitimate. According to the FBI, the operation relied mainly on genuine recordings taken from earlier victims. Investigators also describe synthetic videos of prominent people as an additional technique. A deceptive meeting could therefore show real footage of a real person who had never joined that meeting. FBI campaign overview · FBI advisory

The decisive request came when something supposedly went wrong with the meeting. The people running it offered what looked like a technical fix. The Justice Department says the fix installed malicious software on the participant’s computer. Once that happened, the risk extended beyond the call to the accounts or systems the computer could reach. The familiar face had made what appeared to be ordinary meeting support easier to accept.

The activity authorities describe could also give the attackers material for a future meeting. DOJ says they recorded webcam footage and accessed contact lists and communications accounts. A person brought in through a trusted connection could become the face or apparent sender someone else recognized later. That helps explain why the footage mattered even apart from any cryptocurrency the attackers could reach on that person’s machine. The public reports describe this as a recurring pattern without tracing every victim through every stage. Justice Department announcement

The risk outlasted the call

In several cases described by DOJ, the theft of cryptocurrency and takeover of communications accounts came months or more than a year after the initial compromise. A participant who left a meeting without seeing an immediate loss could still have exposed access that mattered later. The public account gives a range across reported cases. It leaves the timing and route from the initial compromise to individual thefts unresolved. Justice Department announcement

Most business owners will never handle cryptocurrency for a living. They can still recognize the decision in the middle of this incident. A meeting stops working, people are waiting, and someone on the call offers a fix that asks you to change something on your computer. Pause before making the software change. Leave the meeting and contact your own IT team through a channel you already use. If there is a real meeting problem, the people who manage your computer can tell you what to do. If you already ran the fix, tell them what you ran and when, even if everything seems to work afterward.

Permalink

How Work Keeps Moving When Email Goes Down

In this fictional example, a customer calls to move a delivery from Tuesday to Thursday. The person who answers can take the call. Can they also reach dispatch, find out whether Thursday is open and get the new date back to the customer while the inbox is down?

On November 12, Microsoft reported that users had trouble connecting to Exchange Online mailboxes. Its updates said core Teams messaging and online meetings were largely functional during the broader incident, while some Exchange-dependent tasks, including meeting scheduling, could fail. Microsoft reported an earlier, separate Teams interruption that day. That distinction matters if you’re trying to keep work moving. Try the handoff in the existing chat and ask dispatch to confirm that the message arrived. Microsoft incident reports reproduced by Penn

Follow the request through the business

Take the delivery change as an example. The customer already has a number they can call. The person answering needs a place to hand off the request and a way to see whether dispatch picked it up. If the usual team chat works, put the order number, requested date and person responsible for the callback in one existing channel. Have dispatch reply there once it has checked the route. If the route is full, the caller can go back to the customer with a real choice instead of a promise someone else will have to unwind.

If chat is down too, use the phone route you’ve already agreed on. For a small team that might be the dispatch desk first, then the person covering it. Give the person who takes the request ownership of the callback until someone explicitly accepts it. A voicemail can leave a routine request pending. Whoever took the customer’s call still owns the callback until dispatch accepts the handoff. Confirm a new date only after dispatch checks the route. If today’s driver is already on the road, you need someone to answer and confirm what the driver should do.

Outside the company, the best channel may be different. Your staff might share a chat channel, but a supplier or customer may never see it. If your office phones run through the same internet connection as everything else, try the fallback from a phone on cellular service during a drill. A number is useful only if someone can dial it and someone else knows to answer.

Decide where the agreement goes

In ordinary work, the email thread often holds the history. During an outage, a chat message and two phone calls can leave different people with different versions of the same delivery date. If the order system works, have the person confirming the new date update it while the customer is still on the phone. If it’s unavailable, use one designated temporary log with access limited to the people handling the work. Record the order number, the confirmed delivery date, who confirmed it and whether the customer agreed. Copy the decision into the order record when it comes back and mark the temporary entry as transferred. That gives the next shift somewhere to look before it calls the customer or sends the truck.

The Department of Energy’s September 2026 communications drill asked participants to test alternate routes with their incident-response partners. A small business can borrow the test without borrowing the utility sector’s plan. Department of Energy drill guide

Run a fictional delivery change with email closed. Tell everyone it’s a drill and keep test changes out of live orders. When dispatch confirms Thursday, the person who took the call should be able to tell the customer the new date and show the next shift where it was recorded. If the date exists only in a voicemail or someone’s chat history, the next shift still has to figure out what to put on the truck.

Permalink

Will the Team Still Have the Files When Someone Leaves?

A project spreadsheet can be shared with the whole team and still live in one employee’s work OneDrive. Everyone may open it from the same link for months. When that employee leaves, the link can keep working for a while, even though nobody has moved the spreadsheet to a place the team manages.

Microsoft starts the OneDrive cleanup clock when the user account is deleted. The default retention period is 30 days, and organization settings or holds can change what happens. Blocking sign-in does not start that cleanup clock. When the OneDrive goes into the site collection recycle bin, coworkers lose access to the shared files. Microsoft’s OneDrive retention guidance

Block sign-in promptly. When possible, finish the file transfer and access check before deleting the account. The manager names the documents the team still needs. Authorized IT staff can transfer those files to the appropriate team SharePoint library. Then have someone who needs the files open them from the new location using their own account, and replace the old links in shared instructions. Microsoft’s guidance on keeping a former employee’s files

Permalink

The Bog You Can See From Space

Open NASA’s Landsat image of northern Scotland and look for the dark patches of water scattered across the broad uplands. This is the Flow Country, where peat covers much of the landscape. In some places the peat is more than eight metres deep.

A blanket bog can spread over gentle slopes as well as flatter ground. In a cool, wet climate, moss and other plants grow on a surface that stays wet enough to slow decomposition. The remains build up as peat. Over thousands of years, the ground itself grows thicker.

The image shows how far the bog stretches. The peat beneath it shows how long it took to form. The peat holds a record of plants that grew there long before anyone could see the whole landscape from above. UNESCO dates the start of peat accumulation in the Flow Country to about 9,000 years ago. Read the Flow Country World Heritage listing.

Permalink

Microsoft’s Outage Reached Beyond the Inbox

On Thursday, some people who couldn’t open Outlook also couldn’t use Claude to look up a message or check a meeting. Anthropic traced those failures to the Microsoft 365 connector that lets Claude work with mail, calendars and search. Claude for Outlook was affected too. Anthropic’s incident report

Microsoft listed Exchange Online, Teams and SharePoint Online among the services that could be affected. Its updates described servers that failed health checks and steps to restore capacity, including returning servers to service and shifting traffic. Microsoft reported limited recovery earlier in the evening. In its 9:54 p.m. EST update, it said the attempted improvement was falling short and it was pursuing other ways to restore service. Microsoft’s incident updates, mirrored by the University of Pennsylvania

Microsoft was reporting a service outage. Its updates Thursday evening gave no indication of a cybersecurity incident. The root cause remained unconfirmed.

The Claude example shows why a second app can still leave you with the same problem. If both apps need to reach the same mailbox, opening the other one may get you no closer to the information you need. For something time sensitive, like getting a message to the people due on a customer call, keep a current phone list somewhere staff can reach when Microsoft 365 is down.

Permalink

Who Gets Paid When IT Needs Fewer People?

A service can need fewer people without becoming cheaper for the buyer. If AI reduces the labor behind routine IT work, the saving has at least three possible destinations: a lower invoice, more coverage or margin for the provider, or a new bill for the platform and integrations that make the automation possible.

That is not a prediction that prices will stay high. OECD research finds that AI can lower operating costs and make smaller teams viable; it can also reduce search and switching costs, which may put real pressure on price and quality. But the same research says the outcome depends on access to data and compute, integration, portability, and the practical ability to change suppliers. OECD’s competition review describes both paths.

There is a second wrinkle. A cheaper model call is not automatically a cheaper operating service. OECD notes that agentic systems can use far more model capacity per task, while broader deployment can require organisational change and complementary investment. Its 2026 AI-markets review treats declining model prices as necessary but insufficient for a lower effective cost.

So the useful renewal question is not, “Will this service use fewer technicians?” It is: what will the buyer pay for the supported work, what integration or exception work remains, and what would it cost to leave? If comparable alternatives can be switched between cheaply, savings may reach the invoice. If they cannot, headcount alone does not settle who gets paid.

Permalink