Microsoft Advances AI Security While Nvidia Secures $500 Billion Orders

The artificial intelligence sector saw a flurry of activity on November 18, 2025, with significant developments in security, market trends, education, and energy demands. Companies are rapidly rolling out solutions to manage and secure AI, while market watchers debate the future of AI investments. In AI security, Radware introduced its LLM Firewall, a new tool designed to protect generative AI applications from prompt injections and data theft in real time. This solution integrates with Radware's Cloud Application Protection Services and helps organizations comply with data regulations like GDPR and HIPAA. Meanwhile, Palo Alto Networks released its cybersecurity predictions for 2026, dubbing it the 'Year of the Defender,' highlighting the critical role of AI-driven defenses against emerging threats such as AI-driven identity attacks and the need to secure AI training data. AvePoint also launched its AgentPulse Command Center, an AI agent registry that helps organizations track, manage, and secure their AI agents, providing visibility into activity, costs, and potential risks. This comes as 75% of organizations using AI reportedly faced a data breach last year. Further bolstering security, Zscaler became an early adoption partner in Microsoft's Entra Agent ID ecosystem, aiming to enhance AI workflow security through Zero Trust standards and identity management. Zacks Investment Research also integrated its equity research and data into Microsoft 365 Copilot with the new Zacks Equity Research Agent, offering AI-powered guidance within the Microsoft ecosystem. On the market front, major U.S. indexes experienced a downturn as investors sold off technology stocks, putting pressure on AI-related shares. Despite this, some analysts still anticipate a year-end rally. Nvidia CEO Jensen Huang announced substantial orders, totaling $500 billion for 2025 and 2026, signaling robust demand for AI chips. However, the CEO of Hugging Face expressed concerns at Axios BFD, suggesting that the real bubble in artificial intelligence might be specifically within large language models (LLMs), predicting a potential burst next year. He emphasized that AI encompasses much more than just LLMs, pointing to potential in fields like biology, chemistry, images, audio, and video, with Google, Amazon, Nvidia, and IBM continuing to invest heavily across the AI spectrum. Education and training in AI also saw notable advancements. Ferris State University's secure artificial intelligence program received validation from the National Security Agency, making it the first program of its kind to be accredited as a Center of Academic Excellence in Secure Artificial Intelligence. This designation helps students access grants and scholarships for careers in secure and ethical AI. For professionals, INE launched an 'AI in Automation' course to equip network engineers with skills to integrate AI and LLMs into network management, focusing on real-world applications like configuration generation and troubleshooting. In a more unconventional approach, Alpha School, a private institution, uses AI to deliver core subject lessons for about two hours daily, claiming superior academic results compared to national averages, though tuition costs are high, reaching $50,000 per year in Plano, Texas. Additionally, Connecticut's new Small Business Development Center director, Dorian Zwierewicz, plans to prioritize adding training sessions to help small businesses maximize their use of AI. In the financial sector, Circle Asia partnered with Scienaptic AI to enhance credit card decision-making across its banking network, leveraging Scienaptic AI's Business Rule Engine for faster and more accurate approvals, particularly in Southeast Asia, starting with Vietnam. Finally, the growing demand for AI is also putting a spotlight on energy infrastructure. On November 17, 2025, the American Energy + AI Initiative held a summit in Washington D.C., where leaders discussed the urgent need to secure enough firm power, particularly natural gas, to meet the rapidly increasing electricity demands driven by AI growth.

Key Takeaways

  • Radware launched its LLM Firewall on November 18, 2025, providing real-time security for generative AI applications against prompt injections and data theft, while supporting data regulations like GDPR and HIPAA.
  • Palo Alto Networks predicted 2026 as the "Year of the Defender," emphasizing AI-driven defenses against threats like AI-driven identity attacks and the need to secure AI training data.
  • AvePoint introduced its AgentPulse Command Center on November 18, 2025, an AI agent registry for tracking, securing, and managing AI agents, noting that 75% of organizations using AI experienced a data breach last year.
  • Zscaler partnered with Microsoft's Entra Agent ID ecosystem on November 18, 2025, to enhance AI workflow security through Zero Trust standards and identity management.
  • Zacks Investment Research integrated its equity research and data into Microsoft 365 Copilot with the new Zacks Equity Research Agent, offering AI-powered guidance.
  • Nvidia CEO Jensen Huang announced $500 billion in orders for 2025 and 2026, indicating strong future demand for AI chips despite a recent tech stock sell-off.
  • The CEO of Hugging Face warned on November 18, 2025, that a bubble in large language models (LLMs) could burst next year, though he sees significant potential for AI in other fields like biology and chemistry, with Google, Amazon, Nvidia, and IBM investing heavily.
  • Ferris State University's secure artificial intelligence program received NSA validation on November 18, 2025, becoming the first Center of Academic Excellence in Secure Artificial Intelligence.
  • INE launched an "AI in Automation" course on November 18, 2025, to train network engineers in integrating AI and LLMs for tasks like network configuration and troubleshooting.
  • The American Energy + AI Initiative summit on November 17, 2025, highlighted the urgent need for more firm power, specifically natural gas, to meet the rising electricity demands of AI growth.

Radware launches LLM Firewall for AI security

Radware introduced its new LLM Firewall on November 18, 2025, a security solution for generative AI applications. This innovative tool protects large language model prompts from attacks like prompt injections and data theft in real time. It works with all tiers of Radware's Cloud Application Protection Services and helps companies follow data regulations such as GDPR and HIPAA. Constance Stack, Radware's Chief Growth Officer, stated that this solution is crucial for organizations to adopt AI safely. The LLM Firewall is easy to integrate and works with any AI model.

Palo Alto Networks predicts 2026 AI security trends

Palo Alto Networks announced six cybersecurity predictions for 2026 on November 18, 2025, calling it the "Year of the Defender." The company, based in Santa Clara, California, believes AI-driven defenses will become crucial against new threats. Wendi Whitmore, Chief Security Intelligence Officer, explained that AI is changing security risks but also offers opportunities for defenders. Predictions include a rise in AI-driven identity attacks, new insider threats from AI agents, and the need to secure AI training data from poisoning. The company also highlighted executive responsibility for AI risks and the urgent need for post-quantum cryptography.

AvePoint launches AgentPulse for AI agent security

AvePoint introduced its new AgentPulse Command Center on November 18, 2025, in Jersey City, New Jersey. This tool helps organizations track and manage their AI agents to improve security and control costs. AgentPulse, part of the AvePoint Confidence Platform, gives a clear view of AI tools' security status and monitors agent activity. Jeremy Thake, Chief Architect, noted that 75% of organizations using AI faced a data breach last year. The center also helps identify high-activity agents to prevent unexpected charges, transforming AI governance into a business advantage. AgentPulse is currently available to customers in a private preview.

AvePoint unveils AgentPulse for AI agent control

AvePoint launched its new AgentPulse Command Center on November 18, 2025, in Jersey City, New Jersey. This AI agent registry helps organizations track all their AI agents in one place. It boosts security and cost visibility while lowering risks from unmanaged AI agents. Dr. Tianyi Jiang, Chief Technology Officer, stated that AgentPulse empowers customers to use AI agents confidently with necessary controls. Key features include an AI Agent Registry, real-time Security Monitoring, Cost Visibility, and Risk Reduction. AgentPulse is available immediately as part of the AvePoint Confidence Platform.

AvePoint AgentPulse helps secure AI agents

AvePoint announced its AgentPulse Command Center on November 18, 2025, in Jersey City, New Jersey. This AI agent registry, part of the Confidence Platform, helps organizations manage and secure their AI agents. It shows which AI agents are active, who uses them, and what sensitive files they access. Jeremy Thake, Chief Architect, noted that 75% of organizations using AI experienced a data breach last year. AgentPulse also helps control costs by highlighting high-activity agents. John Peluso, Chief Technology Officer, emphasized that AgentPulse turns AI governance into a proactive business enabler. It is currently available to customers in a private preview.

AvePoint introduces AgentPulse for AI agent governance

AvePoint launched its new AgentPulse Command Center on November 18, 2025, in Jersey City, New Jersey. This center enhances AvePoint's AI agent governance tools within the Confidence Platform. It provides more visibility to help reduce security risks and control costs associated with AI agents. Jeremy Thake, Chief Architect, explained that AgentPulse allows IT and security teams to scale AI confidently without compromising data security. John Peluso, Chief Technology Officer, added that it helps organizations avoid unexpected charges from high-activity AI agents. AgentPulse is available to customers in a private preview.

Zscaler and Microsoft Entra boost AI security

Zscaler, Inc. became an early adoption partner in Microsoft's Entra Agent ID ecosystem on November 18, 2025. This partnership aims to improve AI workflows through secure integration and identity management. Zscaler will help ensure every AI agent request is verified under Zero Trust security standards. Raj Judge, EVP of Corporate Strategy, stated this integration empowers customers to bring authorized agents into their AI workflows safely. The collaboration focuses on developing features for agent management and governance, making it easier for organizations to onboard and manage AI agents securely.

Analysts predict year-end rally despite AI pressure

On November 18, 2025, major U.S. indexes fell as investors sold off technology stocks, putting AI under pressure. However, some analysts believe a year-end rally is still likely. Nvidia CEO Jensen Huang announced $500 billion in orders for 2025 and 2026, signaling strong demand for AI chips. Ross Mayfield, a Baird investment strategist, warned that any muted forecast from Nvidia could hurt the market. Meanwhile, Fed officials expressed divided opinions on a December rate cut. Michael Graham, an analyst at Canaccord Genuity, maintains a bullish outlook for a year-end rally.

Year-end rally still possible despite AI slump

On November 18, 2025, major U.S. indexes dropped as investors sold technology stocks, causing an AI slump. Despite this, some analysts, like Michael Graham from Canaccord Genuity, still expect a year-end rally. Nvidia CEO Jensen Huang revealed $500 billion in orders for 2025 and 2026, indicating strong future business. Ross Mayfield, a Baird investment strategist, cautioned that any weak forecast from Nvidia could negatively impact the market. Additionally, Fed officials had mixed views on a December interest rate cut. India's trade deficit also hit a record in October due to gold imports.

Ferris State AI program gets NSA recognition

Ferris State University in Big Rapids, Michigan, announced on November 18, 2025, that its secure artificial intelligence program received validation from the National Security Agency. This makes it the first program of its kind in the nation to be accredited as a Center of Academic Excellence in Secure Artificial Intelligence. The NSA's CAE designation helps schools gain grants and scholarships. Greg Gogolin, who leads the program, stated it prepares students for careers focused on secure and ethical AI use. Students in the bachelor of science program can specialize in areas like cybersecurity and machine learning, using a secured AI lab.

Ferris State AI program earns NSA validation

On November 18, 2025, the National Security Agency validated Ferris State University's secure artificial intelligence program. This Western Michigan school's program is now a Center of Academic Excellence in Secure Artificial Intelligence. This special accreditation helps students get grants and scholarships. Greg Gogolin, who leads the AI program, emphasized its quality in preparing students for careers in secure and ethical AI. Students in the bachelor of science program can focus on cybersecurity, machine learning, and language processing, using a dedicated secured AI lab. Renae Weathers from the NSA noted this program helps address a critical shortage of cybersecurity professionals.

Alpha School uses AI to teach students

Alpha School, a private school in the U.S., uses artificial intelligence to teach students instead of human teachers. Students learn from computer-led lessons for about two hours daily on main subjects like math and reading. Adults called "guides" are present but do not teach or answer questions. The school claims its students achieve better academic results than the national average. Alpha School also focuses on "life skills sessions" for six hours daily, teaching perseverance and self-control. Tuition is high, costing $50,000 per year in Plano, Texas, and $40,000 in Fort Worth, with other campuses in major cities.

Circle Asia partners with Scienaptic AI for credit cards

On November 18, 2025, Circle Asia announced a partnership with Scienaptic AI to improve credit card decisions across its bank network. Scienaptic AI's Business Rule Engine will make credit approvals faster and more accurate. This intelligent platform automates rules and simplifies data, helping Circle Asia grow its lending business efficiently. Arnab Ghosh, CEO of Circle Asia, stated this partnership provides the agile technology needed to scale rapidly. Joydip Gupta, Scienaptic AI's Business Head for APAC, added that their technology will help make credit cards accessible across Southeast Asia, starting with Vietnam.

New director prioritizes AI for Connecticut businesses

Dorian Zwierewicz became the new state director of the Connecticut Small Business Development Center on November 18, 2025. His top priorities include securing funding, promoting international trade, and integrating artificial intelligence for small business growth. Zwierewicz, who previously led an award-winning SBDC in Florida, plans to highlight Connecticut's manufacturing strength globally. He also aims to add training sessions to help small businesses maximize their use of AI. The CTSBDC offers free business advice and helped launch 170 new businesses and create over 550 jobs in 2023.

INE launches AI automation course for engineers

INE, a top provider of network and IT training, released its new "AI in Automation" course on November 18, 2025. This course helps network engineers gain practical experience integrating artificial intelligence and large language models into network automation. Participants will learn to use AI for tasks like generating network configurations, troubleshooting, and enforcing security policies. The curriculum focuses on real-world scenarios to build confidence in using AI tools for network management. This new course aims to equip engineers with skills to create intelligent, efficient, and secure automation workflows.

Zacks brings AI research to Microsoft 365 Copilot

Zacks Investment Research launched the Zacks Equity Research Agent for Microsoft 365 Copilot. This new tool brings Zacks' award-winning equity research and data directly into the Microsoft 365 workflow. It provides AI-powered guidance with strong enterprise-grade controls. This integration allows users to access Zacks' valuable insights more easily within their existing Microsoft 365 environment.

Hugging Face CEO warns of LLM bubble

The CEO of Hugging Face stated at Axios BFD on November 18, 2025, that the real bubble in artificial intelligence might be in large language models, or LLMs. He predicted the LLM bubble could burst next year. However, he sees great potential for AI in other fields like biology, chemistry, images, audio, and video, noting that LLMs are just one part of AI. Hugging Face aims to make good machine learning accessible. Major tech companies like Google, Amazon, Nvidia, and IBM are also investing heavily in AI.

Energy and AI leaders discuss powering AI growth

On November 17, 2025, the American Energy + AI Initiative and its partners held a summit in Washington D.C. to discuss powering AI growth. Leaders from government and industry, including U.S. Secretary of Energy Chris Wright, examined how to quickly add enough firm power for rising AI demand. The Hamm Institute for American Energy has researched how to power AI with reliable energy. Mark P. Mills, executive director of the National Center for Energy Analytics, presented a study on AI's impact on electricity demand. He highlighted the urgent need for more natural gas to meet this demand.

Sources

NOTE:

This news brief was generated using AI technology (including, but not limited to, Google Gemini API, Llama, Grok, and Mistral) from aggregated news articles, with minimal to no human editing/review. It is provided for informational purposes only and may contain inaccuracies or biases. This is not financial, investment, or professional advice. If you have any questions or concerns, please verify all information with the linked original articles in the Sources section below.

AI Security Large Language Models (LLMs) AI Agents AI Governance Cybersecurity Data Security Prompt Injection Zero Trust Security Identity Management Network Automation AI in Education Financial Services AI Small Business AI Energy and AI AI Market AI Investment Generative AI Cloud Security Risk Management Data Regulations Nvidia Microsoft 365 Copilot Palo Alto Networks Radware AvePoint Zscaler Hugging Face Ferris State University IT Training

Comments

Loading...