Become a member of our daily and weekly newsletters for the latest updates and exclusive content about leading AI coverage. Leather
If you pay attention to cyber security news at all, there is a strong chance that you have heard scary reports of companies hiring on external contractors that appear to be hackers or North -Korean spies Off with sensitive, patented data.
But even without that cloak-en-days, international espionage veneer, the truth is that all organizations have reasons to worry about their data security and the prospect of ‘exfiltration“Or the relocation of data without permission. IBMs 2024 costs of a report for data breach Discovered that incidents with data -sex filtration are now in extortion, now on average around $ 5.21 million per incident.
In an era in which data has never been more important or valuable for an organization – but also more than ever between Silos moves – how can companies best protect their sensitive information without breaking the bank?
A new company, Orion Security, believes that generative AI large language models are the key. Today, the company announced its rise by $ 6 million in seed financing under the leadership of PICO Partners and FXP, with the participation of underlining signs VC and prominent cyber security leaders, such as the founders of Perimeter 81 and the Ciso of Elastic.
Orion Security, founded by Nitay Milner (CEO) and Yonatan Kreiner (CTO), is already working together with leading technology companies to help them protect sensitive company data against Insider threats, according to an interview Venturebeat with Milner last week with Milner via Video Call.
“I have spent many years as a product leader in various companies that solve very complicated challenges around perceptibility and security in cloud environments, so that T-Mobile and Blackrock can get their very complex system stacks and understand them better,” said Milner. “I have experienced firsthand that the biggest problem in data security is to understand the business context of how sensitive data is used in a company.”
AI-driven Contextual Data Protection (AI CDP)
In contrast to traditional data protection tools that depend on rigid rules and manual policy, the Platform of Orion Security Dynamic learns the business processes of an organization.
By understanding how data usually goes within an organization, Orion can distinguish between legitimate workflows and potential threats, intentionally or coincidentally.
“Orion makes a revolutionization of data protection by understanding business processes and data flows in the company and automating data loss with the power of AI,” Milner explains.
This approach is a deviation from conventional manual policy -based security models, which, according to Milner, are fundamentally inadequate.
“Most security solutions rely on manual policy, but policy does not scale. There are new applications and workflows that make them outdated quite often.”
He further emphasized how security teams struggle with outdated methods: “Security teams are constantly writing endless policy, and are touched by false positives, and yet data leaks from companies.”
Orion Security uses a combination of its own AI models and refined open-source LLMS to automate data protection.
“All our AI is something we have developed … We do not use a third party, such as Chatgpt or something. We have developed our AI internally, so it is all our IP,” he told Venturebeat.
The platform is based on two core models: one for classification, which identifies how sensitive data is based on context, and another for business reasoning that uses user roles, workflows and typical data movements to detect anomalies.
The AI of Orion is further demonstrated on industry -specific and organizational -specific data to improve accuracy, so that it adapts to the unique activities of each company.
Although they use refined open-source LLMS, Milner notes their surprising effectiveness, even without extensive pre-training, saying: “LLMs who are open source … have a lot of context, and you would not believe the level they give you by simply throwing sensitive data about them.”
How the Orion solution works
The platform connects to the cloud services, browsers and devices of an organization to fully map data flows.
The core of his detection possibilities are the indicators for leakage (IOL) engine, which use own reasoning models and classification of the Great Language Model (LLM) to analyze data movement patterns.
The most important characteristics include:
- Real-time risk assessment: The platform is continuously evaluating business processes, assigning risk scores based on observed behavior.
- Sensitive data detection: Orion identifies and classifies data types, including personally identifiable information (PII), trade secrets, wage data and intellectual property (IP).
- Minimum manual configuration: In contrast to traditional DLP tools that require an extensive installation, Orion automates detection and response with minimal user intervention.
- Reduced false warnings: By including the business context, Orion ensures that security teams are only warned of real suspicious activities, which shortens noise and unnecessary studies.
Milner compares Orion’s approach with solutions for end point detection and response (EDR), but for data protection. “We act as an EDR for data thinking on a crowdstrike for your data.
In addition to catching malicious insiders, Orion also makes a distinction between human errors and external attackers. “The three most important vectors for data leaks are malignant insiders, human errors and external attackers.
Enterprise leaders can see the electricity of their company’s data at a glance
Orion Security offers users a dashboard-driven experience and offers real-time insights into business data flows. The interface categorizes the risk due to severity, allowing security teams to quickly identify and tackle risky activities.
Some remarkable elements of the user interface of Orion include:
- Top data types monitored: The system classifies and follows PII, marketing material, product -related data and source code.
- Distribution of risk score: A visual breakdown of critical, high, average and low risk activities helps to prioritize security reactions.
- Top outdoor sources: Shows the most common platforms where data is transferred, so that security teams detect unusual exfiltration patterns.
- Business Flow Risk Scores: Each monitored business process is assigned a risk score, with specific activities (for example, “engineering teams that move data before they leave the company”) marked on the basis of the seriousness.
This intuitive approach to data security enables security teams to quickly assess potential threats and take immediate action when necessary.
Milner described the visibility options of the platform: “Imagine that you have a dynamic map of all sensitive data movements in your company – between people, devices and applications – and ensuring that it does not leave your organization.”
High investor confidence
Supporting cyber security veterans further strengthens the approach of Orion. Gil Zimmermann, partner at FXP, who previously set up CloudLock (taken over by Cisco), regards the technology of Orion as a long -awaited evolution in data protection:
“AI creates a turning point for data protection, and Orion Security is paramount in this transformation,” he wrote in a prepared statement in a press release to Venturebeat. “The AI approach of Orion dissolves the core challenges that we have had to deal with for years the lack of business context and overwhelming manual work.
In addition to detection, Orion offers flexibility in response mechanisms, so that companies can adjust their approach.
“Some companies want us to block exfiltration data in real -time, while others prefer to receive only reports or train employees about security policy.
What is the next step for Orion Security and its technology?
Orion Security is already working with leading technology companies (confidential because of business agreements) and plans to further refine its AI models to stay ahead of the evolving insider threats.
The onboarding process of the company ensures that customers see immediate value. “We take three months of historical data when on board a new customer, so our AI delivers value from the first day,” explains Milner.
In addition, Orion emphasizes the emphasized privacy-first security architecture. “We do not store sensitive data – only metadata.
With an AI-driven approach that reduces manual workload, false positives and security-blind spots, Orion Security is well positioned to shape the next generation solutions for context conscious data protection.
Source link
