Salt Security

Securing the Agentic AI action layer: protecting the APIs, MCP servers, and agents powering modern digital business.

[Salt Security @SaltSecurity]

@cyberdefensemag 💜🎉🥳

[CyberDefenseMagazine @cyberdefensemag]

🏆Huge congratulations to #SaltSecurity on winning Top Infosec Innovators at Cyber Defense Con 2026! Your bold vision, relentless innovation, and game-changing impact are redefining cyber resilience. Truly well-deserved. The industry is stronger with you! #CDCon2026 #CDMAwards

[Salt Security @SaltSecurity]

We’re very excited to announce that Salt Security has been named a winner of the Coveted Top InfoSec Innovator Awards for 2026!

We’ve been recognized across three categories in @cyberdefensemag 's 14th Annual InfoSec Awards at CyberDefenseCon 2026:
🏅 Most Promising Agentic AI Security Solution
🏅 Trailblazing AI-based Cybersecurity Innovation
🏅 Transformational API Security Solution

See the entire list: https://t.co/u7oxaJhHPo
Read the full story: https://t.co/HmDSyg31TY

Behind every milestone is a team of diligent people pushing the boundaries of what’s possible. Thank you so much to our incredible employees, customers, and partners 💜

[Salt Security @SaltSecurity]

We’ve got an informative webinar coming up, are you tuning in?

On November 4th, our speakers will be looking into what recent incidents mean for keeping organisations safe.
Topics:
· Agents aren’t just chatbots with plugins.
· How AI attacks are already playing out
· Why your current stack is blind and noisy
· What security must look like for agents

Sign up today > https://t.co/2hRb9ovnsc

[Salt Security @SaltSecurity]

@PureAINews Thanks for sharing!

[Pure AI @PureAINews]

Nearly 50% of respondents in a @SaltSecurity study reported suspected or confirmed unexpected AI agent activity, yet only 17.5% could automatically stop misbehaving agents.

Explore the risks behind these findings 👇

#AgenticAI #Cybersecurity #AISecurity
https://t.co/GYirEN7AEv

[The Security Transformation Research Foundation @Transform_Sec]

Nearly Half of Organisations Report Unauthorised #AI Agent Actions, Research Finds

https://t.co/rqPejzdbwk

@IT_SecGuru @SaltSecurity #tech #cybersecurity #business #leadership #management #governance #CISO #CIO #CTO #CEO #cyberthreats #cyberattacks #databreaches #AgenticAI

[JC Gaillard @Corix_JC]

Nearly Half of Organisations Report Unauthorised #AI Agent Actions, Research Finds

https://t.co/O9bsWR1ZS9

@IT_SecGuru @SaltSecurity #tech #cybersecurity #business #leadership #management #governance #CISO #CIO #CTO #CEO #cyberthreats #cyberattacks #databreaches #AgenticAI #AIrisks

[Salt Security @SaltSecurity]

The data is in. Are you ready?
Our newly released State of Agentic AI and API Security Report explores the risks, realities, and research shaping the next era of cybersecurity.

Explore the report 🔍 → https://t.co/fayH5RUIHG

[Salt Security @SaltSecurity]

Are your APIs meeting Australia's CPS 234 security requirements? 🇦🇺

As AI agents become more integrated into business operations, maintaining visibility, security, and compliance is more important than ever.

In our latest blog, we explore how CPS 234 applies to APIs and AI agents, the challenges organizations face, and what it takes to demonstrate effective security controls: https://t.co/eaBAzn6D6u

[Salt Security @SaltSecurity]

Salt's State of Agentic AI and API Security Report is here 🧂

Our latest research dives into the evolving threat landscape, emerging security gaps, and what organizations need to know as AI moves from thinking to acting.

The findings might surprise you 📊

Download the report here > https://t.co/hst8xUp3KR

Read about it in the blog > https://t.co/HyNrWGygWq

[Salt Security @SaltSecurity]

From prompt to action, @cybrrnick breaks down why runtime protection needs to extend across the entire agentic estate.

What that means in practice:
• Full-path visibility
• Prompt-to-API correlation
• Posture context
• Shadow LLM discovery
• One platform, not two vendors

Read the blog for the full story > https://t.co/3nWoAe7huN

[Salt Security @SaltSecurity]

Roey Eliyahu, Salt Security’s CEO and Co-Founder, had a great chat with Raviraj Solanki from IT TechPulse about how AI agents are changing the landscape of API security + the need for visibility across the entire agentic ecosystem.

Check out the entire convo > https://t.co/kDmmDjbLBr

[Devvrat Hans @DevvratHans]

a single email was enough to make an AI agent execute code.

@SaltSecurity found that a malicious email could bypass @ManusAI's prompt-injection defenses using JavaScript obfuscation.

the architecture is the alarming thing here ig:

> email → agent → code execution → connected accounts

the flaw has reportedly been patched, so this is not an active exploit.

but it shows why agent security cannot stop at “did the model detect the prompt injection?”

for people building agents:

every connected tool is part of your attack surface.

https://t.co/K4v6lrUUVA

[קודקוד סייבר @kodkodcyber]

מנוס היא אחת מפלטפורמות ה-AI Agents הבולטות בשוק וזכורה בין היתר בעקבות המחלוקת סביב רכישתה על-ידי Meta והמעורבות הסינית בעסקה.

מחקר חדש של @SaltSecurity מציג כיצד הפלטפורמה הייתה פגיעה לחולשת Indirect Prompt Injection שאפשרה לתוקף להגיע להרצת קוד.

בואו לשרשור 🧵 >>

[Salt Security @SaltSecurity]

The deployment covers thousands of APIs, continuously increasing, and the full AI agent footprint giving their team real-time visibility into every agent call, MCP server interaction, and API touch point across the business.

This is what responsible AI adoption in financial services looks like. Congratulations to the team at New American Funding!

[Salt Security @SaltSecurity]

We're proud to announce that @NewAmericanTeam has selected Salt Security as its agentic and API security platform💜

As one of the nation's largest direct mortgage lenders, New American Funding handles some of the most sensitive data in financial services and their security team set a clear mandate when leadership accelerated AI adoption: security would be designed in from the start, not bolted on after the fact.

[Salt Security @SaltSecurity]

Jeff Farinich, EVP Technology Services and CISO at New American Funding, puts it well:

"We are deploying AI across the business because it drives real value for our customers and our teams. But AI agents operate differently than traditional software as they make decisions, chain tools together, and call APIs in ways that are hard to predict and harder to audit without the right platform. Salt Security gives us the visibility and control we need to say yes to AI with confidence."

[Salt Security @SaltSecurity]

Our recent integration with @AnthropicAI connects Salt Security with Claude Enterprise’s Compliance API, giving security teams continuous visibility into organization-managed MCP servers (including when they’re added, updated, or removed). No agents, traffic analysis, or manual cataloging required.

See it in action 🎬 https://t.co/lQFPg06JsO

[Salt Security @SaltSecurity]

New Salt Labs research just dropped 🧂
Here’s the rundown:
Salt Labs gave an AI agent one seemingly ordinary email… and then things got interesting 👀

Our researchers found that malicious instructions hidden inside an email could manipulate Manus into executing code and reaching connected services like email, cloud storage, and code repositories.

The kicker? Manus’ security guardrail actually did detect the malicious activity, but only after the code had already run.

The bigger lesson here is that AI agents don’t just read and respond, they ACT across tools, APIs, and connected systems. And when the action comes before the alert, detection alone isn’t enough.

Learn more Salt Labs research > https://t.co/0ECOz3Ul21
Watch the video > https://t.co/1MVgWZjTZd