It’s 8:42 a.m. The security operations center is quiet, save for the hum of server fans and you slurping your morning coffee. Then, your main dashboard flashes red. One lockout ticket turns into 10, then 100. An email hits your box:

We have encrypted your systems. Pay $20m in Bitcoin or we will leak your data.

Ransomware is a highly sophisticated, multimillion-dollar enterprise these days. Industry experts warn that the threat actors behind ransomware operate like modern businesses, and they’ve just put your organization in their crosshairs.

In Block the breach, a high-stakes interactive game built on real-world reporting from IT Brew, you are the final line of defense. Every decision you make—from pulling the plug to bringing in a third-party ransomware negotiator—will dictate whether you neutralize the threat or trigger a total organizational collapse. Will you outsmart the hackers, or will you become another cautionary statistic?

Are you ready to save the company?—CN, EH, NK

If you want the job done right—and that job is to repeatedly edit a document—you might have to do it yourself.

Microsoft Research found that LLMs have a way of changing and corrupting a doc during long workflows.

Using a benchmark they’re calling “Delegate-52,” Microsoft’s team tested 19 models across 52 domains, including coding, accounting, and music notation. The tests involved up to 20 interactions, including repeatedly editing, transforming, and recreating a file. And bad news: even top frontier systems, like Gemini 3.1 Pro, Claude 4.6 Opus, and GPT 5.4, corrupted data about one-quarter of the time, on average.

The test. In its research, Microsoft provided an example of an accounting document from fictional nonprofit “Hack Club.” The researchers split the test ledger into separate files by expense category, then merged it back into one file—while regularly throwing in “distractor” topical documents to simulate a work environment. They computed reconstruction scores after 10 round trips of editing and returning.

How to keep document integrity intact during agentic workflows.—BH

Sorry, tired security pro: You’re likely going to need to patch and fix vulnerabilities faster than ever. Here’s why:

There will be more discovered vulnerabilities, thanks to AI models like Anthropic’s upcoming Claude Mythos Preview that can scan codebases quickly. Anthropic says the LLM has already found “thousands of high-severity vulnerabilities,” and has given select industry partners early access so they can patch their systems, an effort dubbed “Project Glasswing.”

Attackers are getting faster at reverse-engineering exploits from patches. Cybersecurity company Palo Alto’s Unit 42 research found that some attackers are scanning for newly discovered vulnerabilities within 15 minutes of a CVE being announced: “Exploitation attempts often begin before many security teams have even finished reading the vulnerability advisory,” Palo Alto wrote in its 2026 global incident report.

Tips for CISOs looking to get ahead of the patching frenzy.—BH

A crossword book made by the people responsible for your daily dose of trivia and business news.