Sysdig

10/20/2026

Sysdig Accelerate ‘26 Japan is a wrap. 🙌

Thank you to our speakers, customers, and partners who made the day what it was. The energy in the room was real, and it came from people who are genuinely serious about defending their cloud and AI environments.

AI is changing how attackers operate. It’s also changing how defenders respond. Keynotes, live demos, real-world customer stories, and panel discussions covered all of it, including how to detect threats at runtime and act at machine speed before the damage is done.

To our customers in Japan: thank you for the trust you place in us. We’ll see you next year. 💚

10/15/2026

Frontier AI models like Mythos can now discover vulnerabilities faster than any human. Security teams are on high alert. And most of the advice floating around isn’t helping. 🔒

Here’s what actually matters: finding more vulnerabilities doesn’t mean you have to fix more. Runtime context shows which ones actually add business risk. AI helps you apply that focus at scale.

Join us on October 29 to separate what’s genuinely changed from the noise.

Mythos, Decoded: What’s Actually Changed in Vulnerability Management

📅 October 29 | 8am PT / 11am ET

You’ll walk away knowing:
🔹 What Mythos actually is, and isn’t, for your program
🔹 Why vulnerability management has plateaued even with mature tools
🔹 Practical steps to keep pace with AI-accelerated discovery without the panic

Register now 👇

https://okt.to/N1vr29

10/14/2026

Today, we’re excited to announce Teri Harrell as Sysdig’s new Chief Human Resources Officer. 🎉

Teri is a proven people leader with more than 25 years of experience in talent acquisition, talent development, and organizational change management. Most recently, she served as CHRO of VitalEdge Technologies, and before that, she held HR leadership roles at Red Hat, where she helped scale one of the world’s largest open source organizations.

Teri will lead the people strategy behind our mission: finding, growing, and keeping the talent that helps our customers defend their AI and cloud environments.

Welcome to the team, Teri!

↳ Read the press release: https://okt.to/JDvalG

10/13/2026

It’s Cybersecurity Awareness Month. Which means it’s time to talk about the things that cause breaches every other month too. 🔒

The data doesn’t lie:

🔹 78% of breaches trace back to system intrusion and social engineering
🔹 Vulnerability exploitation up 55% year over year
🔹 Human element still appears in 62% of breaches
🔹 Global average cost of a breach: nearly $5 million, up 12%

None of this is new. That’s the point.

MFA works. Least privilege works. Patching works. Runtime visibility works. Breaches keep happening because these fundamentals get treated as a once-a-year checklist instead of a daily discipline.

Crystal Morin has put every security fundamental you actually need in one place, with the data behind why each one matters.

Cybersecurity Awareness Month is a fine reminder. Just don’t let it be the only one.

https://okt.to/G8xS3L 👇

10/09/2026

Just wrapped SecTor 2026 in Toronto and a few things stood out. 🍁

AI security messaging was everywhere on the show floor, which validates the category. But what felt different this time: runtime is becoming a real part of the conversation, not just a Sysdig talking point.

People came to us with questions we didn’t have to prompt. Security for AI agents is on everyone’s mind, and the governance angle is resonating more than we expected.

The insight that kept landing: AI agents aren’t reliable narrators about their own actions. You can’t secure what you can only see through the agent’s own account of itself. That’s exactly why multi-layer visibility and correlation matters, and why runtime is the foundation, not an add-on.

We’ve been saying that for a decade. It was good to see the market catching up.

Thanks to everyone who stopped by the booth. The corner location helped, but the conversations made it. 💚

10/07/2026

Most preventive security controls assume that what you inspect before deployment is what will run in production.

That assumption doesn't hold for AI.

In 2018, attackers took nearly a year to weaponize a vulnerability. By the end of 2025, it was hours. Security teams didn't lose their edge. The window to use it did.

We unpacked what that means for defenders, including what an AI-orchestrated attack looks like, why traditional tools fall short, and five practices that hold up in production, in our latest article.

Read it here👇

https://okt.to/Ij8phB

Most preventive security controls assume that what you inspect before deployment is what will run in production. Scan the code.

10/05/2026

September's security news had a theme Crystal Morin doesn't mince words about: a gap between initial access and anyone noticing something was wrong. 🔒

The September briefing is out, and this month's highlights are worth your attention:

🔹 OpenAI research agents accessed a Services Australia Medicare portal, the SEC, Department of Education, and Department of Commerce. Some agents bypassed security controls. The breach wasn't identified for months. OpenAI now warns "dozens" more organizations globally could have been affected.

🔹 A ShinyHunters-linked actor used Claude to scale credential theft across 1.8 million Android apps, verified credentials routed in real time to Telegram.

🔹 ShinyHunters breached Cl0p's leak site, then the FBI, then got arrested, and kept operating anyway.

🔹 The Sysdig TRT profiled a human operator on the marimo CVE who matched AI attack speeds while being more evasive. The lesson: detect the chain shape, not the attacker type.

Crystal's closing question is the one every security team should be sitting with: Runtime is the source of truth. But are you using it, and using it properly?

Read the full September briefing👇

https://okt.to/yClsdt

09/30/2026

SBOMs give you visibility into your software supply chain. But what else do you need to defend against attacks? 🔒

A software bill of materials (SBOM) lists what's inside your software. That helps you find affected systems fast when a new vulnerability is disclosed, but a list can't tell you whether a component is safe. Four things fill the gap:

🔹 Inconsistent tool support. Getting attestations to work across developers and infrastructure teams at scale is painful enough that most organizations don't bother.
🔹 You can't fully trust the SBOM. A developer won't know if their own software contains malware. Like a food label, it tells you the ingredients, not whether they're contaminated.
🔹 Vulnerabilities change. One scan at build time isn't enough. New vulnerabilities emerge daily and a static SBOM goes stale fast.
🔹 Internal threats. The 2024 XZ Utils backdoor was hidden by a malicious contributor who had been providing legitimate code for years. An SBOM can't detect intent.

SBOMs are still worth doing. The world just hasn't caught up to making them reliable enough to trust on their own.

Read the full breakdown👇 https://okt.to/B8QX4f

09/29/2026

A typical cloud attack takes 10 minutes from initial access to impact.

That's not a theory. It's what the Sysdig TRT has measured repeatedly in the wild. And it's why frameworks built for on-premises environments, with a median dwell time of 16 days, don't apply to cloud defense. 🔒

The 555 Benchmark was built around that reality: detect in 5 seconds, triage in 5 minutes, respond in 5 minutes. Not as an aspiration. As the minimum operational tempo required to beat an attacker to the finish line.

Here's what meeting it actually requires:

🔹 Detection signals collected from cloud and security tools within 5 seconds, including visibility into ephemeral assets that may not exist by the time a scan runs
🔹 Full context correlated across all relevant signals within 5 minutes of the first alert, with as little manual effort as possible
🔹 Tactical response initiated within 5 minutes of a high-fidelity detection, using cloud automation to act faster than any human workflow can

This is the benchmark. Download the whitepaper and see how your program measures up. 👇

https://okt.to/23GdFZ

09/27/2026

🎉 We're growing our team in Japan!

Please join us in welcoming Tomoari Yasuda as Sysdig's new Country Manager for Japan.

Japan is one of the fastest-growing markets in cloud and AI adoption, with nearly 1 million developers building software there. Tomoari will lead our regional strategy, sales, technical teams, and partner ecosystem as we help Japanese organizations secure that growth in real time.

The opportunity is significant. So is the threat landscape. Tomoari's here to help customers stay ahead of both. Welcome to the team! 🙌