Cyera

Cyera is the data security platform that empowers security teams to manage and protect all of their company’s sensitive data.

[Cyera @cyera_io]

📈Incident volume remained flat at roughly 27 cases through November 2025 but spiked sharply in December as autonomous coding agents hit production.

The strategic takeaway is that prompts, execution plans, and outputs are sensitive data. The AI interaction layer is the new attack surface.

➡️ Read the full technical methodology and threat report here: https://t.co/qnjBQXc2it

[Cyera @cyera_io]

🤖 Autonomous AI agents represent a measurable category of enterprise risk requiring no external threat actor. @cyera_io Research has formalized and quantified this specific phenomenon under a new designation: Agent-Inflicted Damage.

[Cyera @cyera_io]

The data shows severe operational and financial impacts from unchecked autonomous execution.

Documented failures include the PocketOS production database deletion, a 13 hour AWS outage triggered by troubleshooting agents, and $47,000 cloud bills from infinite execution loops.

[Cyera @cyera_io]

Today, @cyera_io and @Snowflake are partnering to give enterprises full visibility into Snowflake Cortex AI agents and one-click control over the sensitive data they can access.

New integrations allow joint customers to:
➡️ Automatically discover every Cortex AI agent and map its full data blast radius.
➡️ Apply Snowflake sensitivity tags and dynamic masking policies directly from Cyera.
➡️ Query your data security posture in plain language right inside Cortex Analyst.

CISOs and CDO's need to answer the question: how do we move faster on AI without losing control of our data? This is how.

https://t.co/uz2AW4wk9C

[Cyera @cyera_io]

We're coming to @BlackHatEvents....And we're not coming alone. https://t.co/5PBvOVdt3E

[Cyera @cyera_io]

Our very own @yotamseg sat down with @jonfortt on @CNBC's Fortt Knox to talk about where cybersecurity is headed, and why the Mythos moment matters beyond the headlines.

Mythos has elevated the conversation about what agents can do, for good and for bad. The risk is real: autonomous systems have their own identities, behaviors, and paths to sensitive data that most teams have no visibility into. That's the problem @cyera_io solves. https://t.co/NLceqkxZXP

[Cyera @cyera_io]

We are incredibly honored to be named to @Redpoint Ventures’ 2026 InfraRed 100 list.

This recognition highlights the companies building the foundational infrastructure for the AI era. At @cyera_io, we believe that innovation is only as strong as the security supporting it, and we are so proud to be part of the backbone powering this next big wave.

A massive thank you to our incredible team for their endless hard work, and to our amazing customers who trust us on this journey every single day. This milestone belongs to all of you.

🔗 Explore the full list and report: https://t.co/GjtGyjYm2y

[Cyera @cyera_io]

Don’t miss the full episode:

📺 YouTube: https://t.co/qVsSzQ7urg
🍎 Apple Podcasts: https://t.co/ndnFxMzURD
🎧 Spotify: https://t.co/VuGwfIiv9S

[Cyera @cyera_io]

Are we measuring the right things?
Or just the things easy to measure?
This one will sting…

@wadebaker, co-founder of @cyentiainst and the architect of the Verizon DBIR, came on The Watchtower to answer that question with Ash Hunt. The answers are uncomfortable.

Three indictments from the conversation:
• "Metrics are used as weapons more often than they're used to manage a program."
• You can post a stellar MTTR by fixing 1% of your vulnerabilities.
• Security teams think like snipers. CISOs should think like generals.

[Cyera @cyera_io]

The most powerful AI in the enterprise now has real-time data security behind it.

Today, @cyera_io and @AnthropicAI are bringing data security into every Claude Enterprise conversation, deployable in just 30 minutes.

Millions of employees use Claude every day. Security teams finally get the visibility to match:

➡️ Monitor risky activity across every Claude conversation, uploaded file, and user behavior in real time
➡️ Apply the same classification and risk-scoring engine already covering 250+ connectors across your data estate
➡️ Deploy in 30 minutes with no new rulebook and a full audit trail

https://t.co/iYOtLdH2BW

[Cyera @cyera_io]

Cyera has acquired Genie Security to deliver unified AI and data security at the endpoint.

As endpoints become one of the most critical battlegrounds in security, Nadav Noy, Noam Dotan, and the Genie team accelerate our ability to build the AI-powered capabilities this moment demands.

Welcome to the @cyera_io family. Time to build! 🚀

[Cyera @cyera_io]

But none of this starts with us. It starts with the customers who trusted us from day one, who pushed us to rethink what data security could be in the age of AI, and refused to accept the old answers.

They call it disruption. We call it customer obsession.

See the full list: https://t.co/zJbFjRjQFi

[Cyera @cyera_io]

Cyera is #9 on the 2026 @CNBC Disruptor 50 List.

We're in remarkable company: @AnthropicAI, @OpenAI, @MistralAI, and the other organizations defining what the AI era looks like. CNBC put it simply: "@cyera_io is on a roll, redefining how to secure enterprise data in the age of AI."

What got us here:
➡️ Military-grade cybersecurity built for AI
➡️ $1.7B raised from investors who believe in Cyera's mission
➡️ A $9B valuation that reflects the scale of what's at stake

[Cyera @cyera_io]

Securing this new frontier requires moving past traditional endpoint controls and shifting focus to where the real risk lies: mapping, monitoring, and governing the critical data flows these agents can access.

Read the full technical breakdown, attack chain steps, and remediation roadmap from @cyera_io Research here 👉 https://t.co/qrRoqL10y9

[Cyera @cyera_io]

Cyera Research has just disclosed Claw Chain, a series of four chainable vulnerabilities in @openclaw, the open-source platform for autonomous AI agents. At the top of the chain is CVE-2026-44112, a CVSS 9.6 CRITICAL sandbox escape.

[Cyera @cyera_io]

In plain terms, an attacker can weaponize the AI agent to execute the attack chain for them, using the agent's own privileges to achieve:

➡️ Sandbox Escape: Bypass isolation boundaries via filesystem race conditions
➡️ Credential Harvesting: Exfiltrate API keys, tokens, and secrets from environment variables
➡️ Privilege Escalation: Elevate local processes to owner-level runtime control
➡️ Persistent Backdoors: Tamper with configurations to alter future agent behavior

Bottom line: When AI agents are given broad execution power, they can quickly turn into unmonitored insider threats. Because traditional controls see this malicious activity as normal agent behavior, detection is incredibly difficult.

[K logix @Klogix]

Our latest Women in Cyber panel was a great evening focused on emotional intelligence in cybersecurity leadership, meaningful conversations, and new connections. Huge thanks to our speakers, attendees, and sponsors @Cyera_io, @SentinelOne, and Vanta for making it such a success.

[Cyera @cyera_io]

New things are landing at @cyera_io.

We're moving past the era of "more alerts" and into the era of operational context. This month, we're giving your team a unified view of how agents reach and expose sensitive data, the behavioral patterns driving risky data movement, and the tools to enforce retention without disrupting the business.

Watch to see how it all comes together. https://t.co/O4D1pehWhT