Macnica begins selling "DeepKeep," a security platform that identifies risks contained in AI models and learning data and protects AI/LLM
Macnica (Headquarters: Yokohama, Kanagawa Prefecture; Representative Director and President: Kazumasa Hara; hereinafter "Macnica") announced today that it has signed a distributorship agreement with DeepKeep Ltd. (Headquarters: Israel; CEO & Founder: Rony Ohayon; hereinafter "DeepKeep"), a company that promotes AI-native security. This will ensure the security and reliability of AI, and support customers in utilizing AI and achieving business results.
■Accelerating the social implementation of AI and LLM and future challenges
The evolution of AI is affecting a wide range of industries, and in particular in recent years, the emergence of generative AI has improved natural language processing capabilities and automated many tasks. It is expected that half of the tasks currently performed by humans will be automated in the next few decades, and there are high expectations for this. However, as the introduction of AI progresses, various risks are becoming apparent, such as malicious attacks on AI models and unintended information leaks, and how to ensure the security and reliability of AI has become a major issue.
Traditional security focuses on the safety of the entire network and data protection, but AI security also requires ensuring the reliability and fairness of AI models. The risk management and countermeasures for each are different, and different approaches are required in terms of technical expertise and countermeasures.
■ AI-native Security & Trust Platform DeepKeep
DeepKeep's AI-native Security & Trust Platform comprehensively addresses security risks and inference trust issues for AI, LLM, and computer vision models.
Specifically, it has four functions to ensure the safe and secure use of AI.
1. Risk assessment
Conducting Automated Penetration Tests
- Easy setup by simply uploading model data and specifying inspection items
Vulnerability assessment
・Poisoning and backdoor detection
- Prioritize the issues you discover and provide recommendations
2. Prevention
Providing measures to strengthen security against issues
・Pre-processing and post-processing to address identified issues
・AI model repair suggestions
3. Detection
Detection solution that does not impair model performance
- Real-time detection
- Post-event detection
- Detection of abnormal values
4 Mitigation
AI firewall provides mitigation
- Access restrictions upon detection
・Real-time alerts
・Collaboration with the Security Operations Center
<Four functions for safe and secure use of AI>
In addition, discovered risks are displayed with a priority order, allowing the risk management team to quickly consider countermeasures and, if necessary, take immediate action.
Macnica offers solutions in the field of AI TRiSM (Trust, Risk and Security Management) that meet the needs and challenges of our customers by providing a lineup of brighter AI (next-generation anonymization technology) and PrivateAI (personal information anonymization). By adding DeepKeep to these products, we will further advance our efforts to address AI risks and increase the reliability of AI.
[Click here for product details]
https://www.macnica.co.jp/business/dx/manufacturers/deepkeep/
[Click here for product inquiries]
Macnica DeepKeep
TEL: 045-476-2010
Email: deepkeep-biz@macnica.co.jp
Starting today, we are holding the online security conference "Macnica Security Forum 2024 (MSF2024)." DeepKeep will be introduced here, so please be sure to watch it.
[For more information and to apply for MSF2024, click here]
https://msf2024.macnica.co.jp/?pcode=int_pr_deepkeep
【Webinar Summary】
Name: Macnica Security Forum 2024 (MSF2024)
Date: Monday, June 17, 2024 ~ Monday, July 15, 2024
Format: Online *Pre-registration required
Theme: Making Security a Business Power
Participation fee: Free
Organized by: Macnica
*Company names and product names mentioned in this text are trademarks or registered trademarks of Macnica and each company.
*The information published in the news release (including product price, specifications, etc.) is current as of the date of announcement. Please note that the information may be subject to change without prior notice.
About DeepKeep Ltd.
DeepKeep provides safe, fair, accurate, explainable and trustworthy AI security solutions by protecting machine learning pipelines, including vision data models, LLMs and tabular models for risk assessment, prevention, detection, monitoring and mitigation. Only AI-native security built with generative AI can protect diverse source domains, models, datasets and limitless content generation. DeepKeep's enterprise software platform delivers security and reliability throughout the R&D, deployment and product lifecycle of machine learning models.
Company name: DeepKeep, Ltd.
■Established: 2022
■Head office location: Israel
■Representative: Rony Ohay, Founder & CEO
■Website: https://www.deepkeep.ai/
About Macnica
Macnica is Service & Solution Company handles the latest technologies in a comprehensive manner, with semiconductors and cyber security at its core. With operations in 92 locations in 26 countries/regions around the world, the company is leveraging the technical capabilities and global network it has cultivated over its 50-year history to discover, propose, and implement cutting-edge technologies such as AI, IoT, and autonomous driving.
About Macnica: www.macnica.co.jp
Inquiries from the press regarding this matter
Macnica, Inc. https://www.macnica.co.jp
Public Relations Office, Miyahara, Isozaki. E-mail: macpr@macnica.co.jp
Macnica,Inc. 1st Building, 1-6-3 Shin-Yokohama, Kohoku-ku, Yokohama 222-8561