This week brought two warning signs for IT security teams and two big updates from the AI world. A hacking group claims it broke into FBI systems, and old stolen passwords are putting water utilities at risk. At the same time, OpenAI and Anthropic both released new AI models that cost less to run. Here is what IT leaders need to know.
A Hacking Group Claims It Breached the FBI
TechCrunch reported that a hacking group called ShinyHunters says it broke into FBI systems and stole data belonging to agents and job applicants. The claim has not been fully confirmed, but the group has a track record of real breaches at other organizations.
If even a federal law enforcement agency can be targeted, no organization should assume it is too small or too secure to be a target. IT managers should make multi-factor authentication mandatory for every account, not just for senior staff, and should have a clear plan ready for how to respond if a breach claim like this turns out to be true.
Old Stolen Passwords Are Putting Water Systems at Risk
TechCrunch reported that hackers are using passwords stolen in past data breaches to try to break into systems that run parts of America's water supply. Many of these systems still rely on old or reused passwords that were never changed after being exposed elsewhere.
This is a reminder for any IT team managing operational technology, not just office networks, to check whether staff credentials have shown up in known breach lists and to enforce regular password changes. A quick audit of password reuse across your organization is a cheap first step that can close an easy door for attackers.
OpenAI Launches GPT-6 Sol and Luna
TechCrunch reported that OpenAI released two new models, called GPT-6 Sol and Luna, that cost less to run and make fewer mistakes than earlier versions.
Lower cost and fewer errors matter directly for IT budgets. If your team already uses OpenAI's tools, this could mean a lower monthly bill or better results without spending more. Still, test any new model on a small task before moving production work over to it.
Anthropic Cuts Prices with Opus 5.5
TechCrunch reported that Anthropic released Opus 5.5, a new AI model priced lower than before while performing close to its more expensive Fable model.
With two major AI companies cutting prices in the same week, IT leaders comparing AI vendors now have more room to negotiate. This is a good time to review your current AI tool contracts and ask whether a cheaper model can do the job just as well.
This week showed two sides of the tech world at once. Security threats are getting bolder, reaching even the FBI and the water systems people depend on every day, while AI vendors are racing to make their tools cheaper and better. For IT leaders, the lesson is the same either way: keep the basics like passwords and multi-factor authentication tight, and keep reviewing new AI options that could save your team money.