Networking + Security, built to work as One. Cato delivers the converged cloud for the AI era: visibility, context, and control in one architecture.
- Tweets 5,595
- Following 7,551
- Followers 12,648
- Likes 23,719
Meet Cato’s Engine Playground. Test and tune the detectors you’ve built before they go into production- so you know your AI policy actually works.
No more guesswork. Prove your policy works before you publish it.
1
173
Claude just stopped asking which mode you want.
Earlier this week, Anthropic merged Cowork and chat into one Claude experience. Claude can now work out what a task needs: drafting a document, building a deck, reaching into connected systems, and using skills already set up in the account.
In Anthropic’s own example, Claude searches and cross-references a legal research database without the user opening it.
By default, Claude asks before it acts. Users can switch that off and let it keep working, checking in only when something needs a closer look.
For security teams, that changes three things at once. Employees using Claude are now working with an agent. Developers who build on Claude can carry a similar reach. And workflows can keep running when nobody is watching.
Our white paper, Cato AI Security: Protecting Anthropic AI, looks at all three: discovering where Anthropic AI is in use, inspecting prompts and responses inline, keeping sensitive data out of them, and auditing what agents did with the tools they accessed. Learn more: bit.ly/4tPRbJT
1
2
292
Blocking AI tools can feel like the safe answer. It can also slow the business down. For @MoonPay, that wasn’t an acceptable tradeoff.
With more than 35 million customers and 500 partners, the company needed the freedom to move quickly. But its security team still needed to understand where information was going and what was happening with it.
That left Doug Innocenti, CIO/CSO at MoonPay, with a harder question: how do you secure AI usage without getting in the way of growth?
Cato AI Security gave him a way to change the answer—and the confidence to say yes instead of no. It helps apply security controls as employees interact with AI, preventing risky activity while guiding them toward safer ways to use it.
Watch the MoonPay story: bit.ly/4y8PFG2
190
The biggest change AI is bringing to ransomware may be who can run it.
AI is lowering the skill needed to carry out a competent attack, and shrinking the time it takes.
@CatoCTRL’s latest research shows how ransomware’s AI adoption is changing. 🧵
1
1
185
Put it together and three things start to change:
💨 Speed: Days of work can compress into hours.
📈 Scale: AI can take on work across multiple victims.
👥 Who: Less experienced attackers can get help with tasks that once required specialist knowledge.
1
30
The defensive lesson matters just as much.
AI has changed who can run the attack and how quickly — it has not yet changed what stops it.
The fundamentals still matter.
Read the full @CatoCTRL research and what defenders should watch for: catonetworks.com/blog/ransom…
34
AI adoption has outpaced governance. And with the EU AI Act, organizations within its scope have binding requirements for how AI is understood, developed, deployed, and used.
That puts pressure on organizations to see how AI is used across the business, the data it can access, and the controls in place to govern it.
📚 This eBook is the six-step guide for moving from regulatory requirements to secure AI governance: catonetworks.com/resources/n…
203
In racing, every second matters. More than ever, that includes the ones before the car lines up on the starting grid.
Read the full story: bit.ly/4AbsLzf
32
