NVIDIA用Nemo護欄和NIM微服務提高了AI安全性的標準

NVIDIA（NVDA）正在通過引入Nemo護欄系統和創新的NIM微服務來提高人工智能安全性的標準。

Artificial intelligence (AI) is rapidly transforming various industries, offering new possibilities and efficiencies. However, ensuring the safety, security, and compliance of AI systems is crucial for widespread adoption and trust in the technology. Recognizing this need, NVIDIA (NASDAQ:NVDA) has introduced two innovative solutions: the NeMo Guardrails system and Garak, an open-source toolset. These solutions aim to address key challenges in AI application safety and scalability.

人工智能（AI）正在迅速改變各種行業，提供新的可能性和效率。但是，確保AI系統的安全性，安全性和合規性對於廣泛採用和對技術的信任至關重要。認識到這種需求，NVIDIA（NASDAQ：NVDA）引入了兩種創新的解決方案：Nemo Guardrails System和Garak，一個開源工具集。這些解決方案旨在應對AI應用程序安全性和可擴展性中的關鍵挑戰。

At the center of NVIDIA's AI safety initiative is the NeMo Guardrails platform, which provides a comprehensive set of tools to ensure AI systems function within strict safety and compliance parameters. The platform's newly revealed NIM microservices offer developers a robust solution for preventing negative outputs and defending against attempts to bypass security protocols.

Nemo Guardrails平台是NVIDIA的AI安全計劃的中心，該平台提供了一套全面的工具，可確保AI系統在嚴格的安全性和合規參數中的功能。該平台新揭示的NIM微服務為開發人員提供了一種強大的解決方案，可防止負面輸出和防禦試圖繞過安全協議的嘗試。

With sectors like retail, automotive, and healthcare relying heavily on AI, the ability to maintain compliant and secure interactions is paramount. NVIDIA's NeMo Guardrails is designed to meet these needs by implementing various protective features, including content moderation, topic management, and jailbreak detection tools. These features work together to ensure that AI systems remain safe, trustworthy, and reliable.

由於零售，汽車和醫療保健等領域嚴重依賴AI，因此保持合規和安全的互動的能力至關重要。 NVIDIA的NEMO Guardrails旨在通過實施各種保護功能，包括內容審核，主題管理和越獄檢測工具來滿足這些需求。這些功能共同努力，以確保AI系統保持安全，值得信賴和可靠。

Key Features of NeMo Guardrails:

Nemo護欄的主要特徵：

Content Moderation: Leverages the Aegis Content Safety Dataset, which comprises over 35,000 human-annotated instances, to identify and filter harmful content generated by AI systems.

內容適度：利用AEGIS內容安全數據集，該數據集包含35,000多個人類註銷的實例，以識別和過濾AI系統產生的有害內容。

Topic Management: Ensures AI systems adhere to specific topics and prevents them from engaging in conversations or generating outputs outside the defined scope.

主題管理：確保AI系統遵守特定主題，並防止它們參與對話或在定義範圍之外產生輸出。

Jailbreak Detection: Monitors AI systems for attempts to bypass safety protocols or generate outputs that violate established rules and regulations.

越獄檢測：監視AI系統試圖繞過安全協議或產生違反既定規則和法規的產出。

These features are essential for industries where maintaining AI safety is critical, and they are supported by the Aegis Content Safety Dataset, which NVIDIA describes as a library of over 35,000 human-annotated instances. This dataset is available for free to developers, providing them with valuable resources to enhance their AI models' safety.

這些功能對於維持AI安全至關重要的行業至關重要，並且由AEGIS內容安全數據集支持，Nvidia將其描述為超過35,000個人類批准實例的庫。該數據集可免費提供開發人員，為他們提供寶貴的資源，以增強其AI模型的安全性。

Companies Already Leveraging NeMo Guardrails:

公司已經利用Nemo Guardrails：

In retail, Gap (NYSE:GPS) is utilizing NeMo Guardrails to ensure its AI-powered virtual stylist generates safe and appropriate responses while assisting customers.

在零售業中，GAP（NYSE：GPS）正在利用Nemo護欄來確保其AI驅動的虛擬造型師在協助客戶時會產生安全且適當的響應。

For automotive applications, Mercedes-Benz (OTCMKTS:DDAIF) is employing NeMo Guardrails to enhance the safety and compliance of its in-car AI assistant.

對於汽車應用，梅賽德斯 - 奔馳（OTCMKTS：DDAIF）正在使用Nemo護欄來增強其車內AI助手的安全性和遵守情況。

Within the entertainment industry, Netflix (NASDAQ:NFLX) is leveraging NeMo Guardrails to ensure its AI-powered movie and TV show recommendations adhere to specific guidelines and avoid generating harmful outputs.

在娛樂行業中，Netflix（NASDAQ：NFLX）利用Nemo Guardrails確保其AI驅動的電影和電視節目建議遵守特定指南並避免產生有害的產量。

These early adopters demonstrate the versatility and scalability of NeMo Guardrails across a wide range of industries, highlighting the growing demand for secure and trustworthy AI solutions.

這些早期的採用者展示了廣泛行業的Nemo護欄的多功能性和可擴展性，強調了對安全和值得信賴的AI解決方案的需求不斷增長。

Moreover, NVIDIA has also released Garak, an open-source toolkit, to identify flaws in major language models. This toolset enables developers to evaluate the security and integrity of AI systems by detecting potential vulnerabilities and inappropriate outputs.

此外，NVIDIA還發布了開源工具包Garak，以識別主要語言模型中的缺陷。該工具集使開發人員能夠通過檢測潛在漏洞和不適當的輸出來評估AI系統的安全性和完整性。

With the launch of both NeMo Guardrails and Garak, NVIDIA is providing developers with the essential tools to build scalable, secure, and compliant AI systems. As AI continues to shape the future of industries worldwide, ensuring the safety and dependability of these systems is more important than ever. NVIDIA's commitment to advancing AI security and safety is a significant step forward in fostering trust in AI technologies across various sectors.

隨著Nemo Guardrails和Garak的推出，NVIDIA為開發人員提供了建立可擴展，安全和兼容的AI系統的重要工具。隨著AI繼續塑造全球行業的未來，確保這些系統的安全性和可靠性比以往任何時候都重要。 NVIDIA致力於提高AI安全性和安全性是促進對各個部門AI技術的信任的重要一步。

Stay tuned for more updates on the latest AI advancements and how companies like NVIDIA are paving the way for secure and scalable AI solutions.

請繼續關注有關最新AI進步的更多更新，以及像NVIDIA這樣的公司如何為安全可擴展的AI解決方案鋪平道路。