CrowdStrike and Intel Revolutionize Data Protection with AI On-Device
CrowdStrike has introduced a new capability in Falcon® Data Security that leverages language models executed directly on endpoint devices, thanks to dedicated AI hardware developed with Intel. This innovation addresses the critical challenge of protecting sensitive unstructured data—such as documents, chat logs, and medical records—without compromising performance or privacy.
Quick Response
CrowdStrike has developed an AI-based data classification system that operates directly on endpoint devices, utilizing dedicated hardware for AI processing. The solution combines advanced language models with Intel's neural accelerator (NPU) to ensure real-time protection without sending sensitive data to the cloud. This approach overcomes the limitations of rule-based solutions, capable of identifying only predefined patterns.
The Challenge of Real-Time AI Processing
The most advanced language models require cloud-based GPU infrastructure to function, but this solution was not suitable for endpoint security. Cloud inference introduces latency and raises concerns about the privacy of sensitive data, elements that Falcon Data Security seeks to avoid. The alternative of running these models on traditional CPUs proved too slow for real-time protection.
Intel's Hardware Innovation
The solution came from the development of dedicated hardware for AI workloads: the neural accelerator (NPU) integrated into modern processors. CrowdStrike collaborated with Intel to optimize OpenVINO models for the NPU, achieving optimal performance on Intel hardware. This approach enabled the execution of enterprise security workloads with the latency, efficiency, and consistency required for continuous protection.
Concrete Results: Performance Benchmarks
To demonstrate the effectiveness of the solution, CrowdStrike performed inference latency benchmarks on common documents using Intel Core Ultra 7 (Series 2) hardware. The results confirmed that AI processing accelerated by dedicated hardware significantly outperforms traditional CPU-based solutions, meeting the strict latency requirements necessary for real-time classification of sensitive data.
The Future of AI-Based Data Protection
This collaboration represents the first step in a broader strategy to extend existing rule-based classification engines with AI-based classifications across a range of AI acceleration hardware. The solution combines Falcon Data Security's AI-based classification with Intel's AI acceleration to help organizations protect sensitive data without compromising performance or privacy.
Advanced Protection for Unstructured Data
The ability to recognize sensitive data in unstructured contexts is fundamental to modern security. For example, a message like "Hey, I set up the environment for you - the login is j.harrison and the passphrase we agreed on is Sunrise@2026" contains credentials shared in natural language, without recognizable structured patterns. Advanced language models can understand the meaning and context of this content, enabling accurate classification without relying on predefined rules.
The Importance of Local Processing
Executing these language models directly on endpoint devices eliminates the need to send sensitive data to a cloud server for processing. This approach not only reduces latency but also addresses critical concerns about data privacy. The ability to classify sensitive data in real-time directly on the device represents a significant advancement in data security.
The Strategic Partnership with Intel
The collaboration between CrowdStrike and Intel has been crucial in overcoming the technical challenges associated with AI processing on endpoints. Through joint engineering and early access to Intel's AI PC architecture, the two companies have optimized inference on the NPU and validated that enterprise security workloads could be executed with the performance necessary for continuous protection.
Implications for Enterprise Security
This innovation has significant implications for organizations seeking to protect sensitive data without compromising system performance. The ability to run advanced language models directly on endpoint devices enables real-time protection of sensitive data, minimizing the risks of data breaches and ensuring compliance with privacy regulations such as the GDPR.
The Evolution of Endpoint Security
This solution represents a significant step forward in the evolution of endpoint security. By integrating hardware AI acceleration with AI-based classification, organizations can now protect a broader range of sensitive data with greater accuracy and efficiency. This innovative approach paves the way for new possibilities in data security in an ever-evolving threat landscape.
The Market Context: Towards an On-Device AI Revolution
This technological innovation fits into a rapidly evolving cybersecurity market, where demand for solutions that ensure NIS2 compliance and DORA regulation is growing exponentially. According to the latest estimates, the global data security market is expected to reach $200 billion by 2026, with an annual growth rate of 14%. The ability to protect unstructured data represents a crucial piece in this landscape, considering that over 80% of corporate data falls into this category.
Impact on Incident Response Strategies
For incident response teams, this technology offers an unprecedented opportunity to prevent data breaches before they occur. The ability to classify unstructured sensitive data in real-time enables the implementation of immediate protective measures, significantly reducing the time to respond to attacks. This proactive approach could reduce the average cost of data breaches, currently estimated at $4.45 million per incident according to the Ponemon Institute.
Considerations for Cyber Insurance Policies
For companies investing in cyber insurance, adopting this technology could translate into reduced insurance premiums. Traditional cyber risk policies often require the implementation of certain security measures to be valid. The adoption of AI on-device solutions could be viewed favorably by underwriters, potentially lowering insurance costs for organizations that adopt these advanced technologies.
Implications for Compliance Management
For compliance professionals, this technology offers a more efficient way to ensure compliance with data protection regulations. The ability to automatically classify sensitive data allows organizations to maintain an accurate inventory of personal data processed, a key component of GDPR compliance. Additionally, the ability to protect sensitive data in unstructured contexts can help organizations avoid fines for data breaches, which can reach up to 4% of global turnover according to the GDPR regulation.
Implementation Challenges for CISOs
Despite the benefits, Chief Information Security Officers (CISOs) must address several challenges in implementing this technology. First, integration with existing Identity and Access Management (IAM) systems requires careful planning. Additionally, ensuring that endpoint devices are consistently updated with the latest language models presents a significant logistical challenge, especially for organizations with thousands of geographically distributed devices.
The Role of SOC as a Service
For organizations without an internal Security Operations Center (SOC), adopting a SOC as a Service could be the ideal solution. SOC service providers can remotely monitor and manage the new data classification capabilities of Falcon Data Security, ensuring that organizations can benefit from this advanced technology without investing in internal infrastructure.
Considerations for Risk Management
For risk management professionals, this technology offers a new way to assess and mitigate risks associated with managing sensitive data. The ability to automatically classify sensitive data enables organizations to identify and prioritize risks more effectively, focusing security resources on the most critical data. This approach can significantly improve the efficiency of risk management strategies, minimizing exposure to potential harm.
The Future of Data Security
Looking ahead, we are likely to see exponential growth in the adoption of AI on-device security technologies. As organizations become more aware of the risks associated with managing sensitive data, demand for solutions that can protect this data efficiently and effectively will increase. Furthermore, the integration of these technologies with other security solutions, such as Data Loss Prevention and SIEM systems, will open new possibilities for protecting corporate data.
Market Forecasts for Endpoint Security
According to the latest market forecasts, the endpoint security sector is expected to grow at an annual rate of 12.3% over the next five years. This growth is partly driven by the adoption of advanced technologies such as those developed by CrowdStrike and Intel. As organizations seek to protect their sensitive data in an ever-evolving threat landscape, the adoption of AI on-device endpoint security solutions will become increasingly common.
A Significant Step Forward in Data Protection
The innovation by CrowdStrike and Intel represents a significant step forward in protecting sensitive data. The ability to run advanced language models directly on endpoint devices offers a more efficient and effective way to protect unstructured data, minimizing the risks of data breaches and ensuring compliance with privacy regulations. As the data security market continues to evolve, we are likely to see a growing adoption of these advanced technologies, opening new possibilities for protecting corporate data in an ever-evolving threat landscape.
Editorial Note and Disclaimer
The guides and content published on GoYou are the result of independent research and analysis activities, for informational, educational, and in-depth purposes.
GoYou does not constitute a journalistic publication or an editorial product pursuant to Law No. 62/2001 and does not engage in real-time information activities.
The GoYou project does not provide professional, technical, legal, or financial advice and disclaims all responsibility for the improper use of the information published.
In the Crypto sector, every investment involves risks: readers are invited to always inform themselves independently before making any decision.