amazon, nvidia and apple Updates

The artificial intelligence landscape is rapidly evolving with new security solutions, educational initiatives, and governmental considerations. Companies like Fabrix Security have launched AI-native identity security solutions, backed by $8 million in seed funding from investors including Norwest Venture Partners, to manage and secure both human and non-human identities. Astrix Security and BeyondTrust are also addressing AI agent security, with Astrix introducing an AI Agent Control Plane (ACP) for secure deployment and BeyondTrust offering new AI security controls for visibility and action. These tools focus on least privilege principles and just-in-time access to reduce risks. Meanwhile, Amazon Web Services (AWS) is optimizing AI training with SageMaker HyperPod's topology-aware scheduling, aiming for greater efficiency. However, security vulnerabilities remain a concern, with thousands of AI servers, including those using NVIDIA components, at risk due to unpatched or exposed systems, as highlighted by Trend Micro. On the educational front, IIT Kharagpur is making AI and Data Science degrees more accessible by offering them in Bengali, aligning with the National Education Policy 2020. In the public sector, Ohio has launched an AI-powered chatbot app, Safeguard Ohio, to help citizens report suspicious activity related to crime and terrorism, reportedly tripling tip submissions. On the policy side, a Trump administration adviser has emphasized the need for the U.S. to lead in AI development, advocating for minimal government intervention to foster private sector innovation. Wearable Devices Ltd. is also contributing to the AI interaction space with its AI-powered neural input technology enabling touchless control for devices like the Apple Watch. In a move to broaden access to AI solutions, Inception has made seven of its AI-native products available on the Microsoft Azure Marketplace as part of a partnership with Microsoft.

Key Takeaways

• Fabrix Security has secured $8 million in seed funding for its AI-Native Identity Security solution, which manages human and non-human identities.
• Astrix Security and BeyondTrust are introducing new tools to enhance the security of AI agents, focusing on secure deployment and access controls.
• Thousands of AI servers are at risk from cyberattacks due to vulnerabilities in components like ChromaDB, Redis, and NVIDIA Triton, and accidental internet exposure.
• IIT Kharagpur is now offering Bachelor of Science degrees in AI and Data Science in Bengali to increase accessibility.
• Ohio has launched an AI chatbot app, Safeguard Ohio, to aid citizens in reporting crime and terrorism-related suspicious activity, leading to a reported tripling of tips.
• A Trump administration adviser advocates for U.S. leadership in AI development with minimal government regulation to encourage private sector innovation.
• Wearable Devices Ltd. is developing AI-powered neural input technology for touchless control of devices, including the Apple Watch.
• Inception has made seven AI-native products available on the Microsoft Azure Marketplace through a partnership with Microsoft.
• AWS SageMaker HyperPod has been enhanced with topology-aware scheduling to optimize AI training efficiency.
• Webinars are being hosted to address AI security challenges, focusing on data governance, privacy, and enterprise-grade AI security standards.

Fabrix Security launches AI-native identity security with $8M funding

Fabrix Security has launched its new AI-Native Identity Security solution, designed to help companies manage and secure both human and non-human identities. The company announced it has raised $8 million in seed funding, led by Norwest Venture Partners, Merlin Ventures, and G Squared Capital. This funding will support further product development and expand sales and marketing efforts. Fabrix Security aims to automate identity and access management tasks, enforce least privilege, and reduce the attack surface for enterprises. The solution uses AI agents to manage identities across cloud, on-premise, and SaaS environments.

Astrix Security introduces AI Agent Control Plane for secure AI deployment

Astrix Security has launched its AI Agent Control Plane (ACP), the first solution designed for secure AI agent deployment. ACP provides AI agents with short-lived credentials and just-in-time access based on least privilege principles, reducing risk and improving auditability. This new tool complements Astrix's existing platform, which helps discover and secure AI agents and non-human identities (NHIs). The ACP aims to enable enterprises to deploy AI agents responsibly at scale, enhancing developer productivity and security oversight.

BeyondTrust offers AI agent security controls for visibility and action

BeyondTrust has released new AI security controls for its Identity Security Insights and Pathfinder Platform. These features provide visibility into AI agents, secure orchestration of their actions, and an AI decision-support layer. The new capabilities include AI Agent Insights for discovering and risk-scoring agents, Model Context Protocol (MCP) orchestration for secure action brokering, and an AI decision-support layer for real-time guidance. BeyondTrust aims to help organizations manage the risks associated with AI agents and integrate them securely into their operations.

Astrix Security launches AI Agent Control Plane for secure AI deployment

Astrix Security has introduced the AI Agent Control Plane (ACP), the first solution designed for secure AI agent deployment across enterprises. ACP ensures AI agents receive short-lived, precisely scoped credentials and just-in-time access, adhering to least privilege principles. This launch completes Astrix's Discover-Secure-Deploy framework for AI agent security. The ACP aims to improve auditability, secure agent access, and boost developer productivity by allowing pre-defined policies for agent use.

IIT Kharagpur to offer AI and Data Science degrees in Bengali

IIT Kharagpur will now offer its four-year Bachelor of Science degrees in Artificial Intelligence (AI) and Data Science in the Bengali language. This initiative aligns with the National Education Policy 2020 and aims to make these contemporary courses accessible to students from India, Bangladesh, and around the world. Director Suman Chakraborty stated that BS degrees, equivalent to BTech and BSc, will be awarded. Short courses for diplomas and certificates will also be available in Bengali, though technical aspects will still be taught in English.

IIT Kharagpur offers AI and Data Science degrees in Bengali

IIT Kharagpur is launching its Bachelor of Science degrees in Artificial Intelligence (AI) and Data Science in Bengali, following the National Education Policy 2020. This move aims to benefit students from India, Bangladesh, and globally who speak Bengali. Director Suman Chakraborty mentioned that these four-year BS degrees are equivalent to BTech and BSc. The institute will also offer diploma and certificate courses in Bengali. While core technical content will remain in English, the translation and teaching will be handled by proficient Bengali-speaking professors.

Ohio launches AI chatbot app to help fight crime and terrorism

Ohio has launched Safeguard Ohio, an AI-powered chatbot app designed to help citizens report suspicious activity related to crime and terrorism. The multi-lingual app allows users to submit information, including video, audio, and photos, which the AI then processes and uses to ask follow-up questions. Reports can be categorized into eight areas, such as drug activity, human trafficking, and school threats. The app has reportedly tripled the number of tips received since its launch on August 6, 2025, and can be used anonymously. It is not a substitute for 911 for emergencies.

Inception offers AI products on Microsoft Azure Marketplace

Inception, a G42 company, has made seven of its AI-native products available on the Microsoft Azure Marketplace. This expansion is part of a strategic partnership between G42 and Microsoft, aiming to provide global organizations with enterprise-grade AI solutions. The available products include tools for executives, insight generation, intelligence analysis, procurement, business process automation, productivity enhancement, and customer experience. This move allows businesses worldwide to integrate AI at scale and accelerate their AI initiatives.

AWS SageMaker HyperPod optimizes AI training with topology awareness

Amazon Web Services (AWS) has enhanced its SageMaker HyperPod platform with topology-aware workload scheduling for AI training. This feature optimizes distributed training tasks across clusters of accelerators by considering network topology to minimize latency and improve efficiency. The system intelligently schedules tasks involving heavy data exchange on nodes with minimal network hops. This enhancement aims to reduce communication overhead and speed up the training of large AI models, potentially leading to significant efficiency gains and faster iteration cycles.

Thousands of AI servers at risk due to security vulnerabilities

A new report from Trend Micro reveals that thousands of AI servers are at risk from cyberattacks due to unprotected or unpatched components. Key vulnerabilities have been found in components like ChromaDB, Redis, and NVIDIA Triton. The research also highlights accidental internet exposure of AI servers, with over 10,000 Ollama servers found exposed without authentication. Trend Micro urges IT professionals to implement secure AI application stacks, including improved patch management, vulnerability scanning, and configuration controls, to prevent data theft and other attacks.

Trump adviser: US should lead AI race, not California

Sriram Krishnan, a senior adviser to Donald Trump on artificial intelligence, stated that the Trump administration does not want California to dictate the future of AI regulation nationwide. Krishnan framed the AI race as an existential competition with China and advocated for minimal government intervention, allowing the private sector to innovate freely. He emphasized that the administration's AI Action Plan aims to keep the U.S. ahead in AI development while prioritizing American workers.

Wearable Devices launches AI touchless control for Apple Watch

Wearable Devices Ltd. has announced its AI-powered neural input technology is being featured in an editorial by NetworkNewsWire. The company's products, the Mudra Band for Apple Watch and the Mudra Link platform, enable touchless, gesture-based control of digital devices. This technology aims to provide more natural and seamless interaction compared to traditional touch screens and voice assistants. Wearable Devices is focusing on the growing demand for AI-driven, touchless wearables in consumer electronics, augmented reality, and enterprise computing.

Webinar to cover AI security challenges and solutions

Optiv is hosting a live webinar titled 'Navigating AI Security: Overcoming Market Challenges and Client Pain Points.' The webinar is designed for senior cybersecurity professionals seeking to understand and mitigate risks associated with AI integration. Participants will learn about specialized advisory and technology services for data governance, privacy, and enterprise-grade AI security standards. The session will explore strategies for responsible AI adoption, secure product development, and compliance with data privacy policies.

Sources

• Fabrix Security Emerges from Stealth with $8M Seed Funding to Launch AI-Native Identity Security Solution, Transforming Enterprise Identity Security
• Securing the Agentic Era: Introducing Astrix's AI Agent Control Plane
• BeyondTrust Delivers Identity Security Controls for AI, Turning Agent Visibility into Action
• Astrix Security Announces the First AI Agent Control Plane (ACP) for Secure-by-Design Agent Deployment
• IIT Kharagpur to teach BS courses in AI and Data Science in Bengali
• IIT Kharagpur to teach BS courses in AI and Data Science in Bengali
• Ohio brings on artificial intelligence chatbot app to help fight crime, terrorism
• PRESSR: Inception expands global reach with AI-native offerings on Microsoft’s Azure Marketplace
• AWS Enhances SageMaker HyperPod with Topology-Aware AI Training Optimization
• Security alert for AI servers: thousands are at risk
• 'We don't want California to set the rules for AI across the country,' Trump adviser says
• AI-Powered Neural Input Wearables: Wearable Devices Ltd. Launches Touchless Control for Apple Watch
• Live Webinar | Navigating AI Security: Overcoming Market Challenges and Client Pain Points