NSFOCUS

06/10/2026

Between late May and early June, several high-profile Instagram accounts were reportedly taken over. The attackers neither exploited software vulnerabilities, nor bypassed authentication mechanisms. Instead, they successfully gained control of the targeted accounts simply by sending a request to AI support assistant.

🤔 More concerningly, some victims reported that after their accounts were compromised, they were unable to reach human support. The AI support assistant had become the sole support channel, with no effective escalation or appeal mechanism available.

🤖 The AI is incapable of verifying a user’s real identity, yet it was granted permissions to perform high-privilege actions. The core issue exposed is that the AI agent was granted ex*****on privileges while simultaneously being entrusted with determining authorization boundaries.

🔗 Read our recap: https://nsfocusglobal.com/ai-security-incident-case-account-takeover-due-to-meta-ai-support-assistant-authorization-flaw/

Overview Between late May and early June 2026, several high-profile Instagram accounts were reportedly taken over by attackers, including Barack Obama’s White House account, the personal account of U.S. Space Force Chief Master Sergeant Bentivegna, and the official account of beauty brand Sephora....

06/09/2026

A core risk within AI security threats lies in the 𝙧𝙚𝙡𝙞𝙖𝙗𝙞𝙡𝙞𝙩𝙮 of AI models, manifested as distorted outputs, hallucinations, and the generation of misleading content.

A cemetery photograph taken after an airstrike hitting a school in Iran was submitted to and for authenticity verification. Controversy emerged when both models produced confident but entirely incorrect assessments.

🧐 This case exposes more than just 𝘼𝙄 𝙝𝙖𝙡𝙡𝙪𝙘𝙞𝙣𝙖𝙩𝙞𝙤𝙣𝙨. The real danger stems from the combination of hallucinations and the authoritative tone adopted by AI outputs.

Read our recap: 🔗 https://nsfocusglobal.com/ai-security-incident-case-both-grok-and-gemini-hallucinated-when-verifying-minab-cemetery-photo/

Overview A core risk within AI security threats lies in the reliability of AI models, manifested as distorted outputs, hallucinations, and the generation of misleading content. While these issues may seem like mere technical flaws, they have evolved into tangible harms in real-world information ecos...

06/05/2026

In the report 𝙉𝙖𝙫𝙞𝙜𝙖𝙩𝙚 𝙏𝙝𝙚 𝘼𝙄-𝙉𝙖𝙩𝙞𝙫𝙚 𝘾𝙡𝙤𝙪𝙙 𝙎𝙚𝙘𝙪𝙧𝙞𝙩𝙮 𝙀𝙘𝙤𝙨𝙮𝙨𝙩𝙚𝙢 𝙄𝙣 𝘾𝙝𝙞𝙣𝙖 recently released by Forrester, NSFOCUS was interviewed as a representative vendor. 🔗 https://lnkd.in/gveNBi3F

Enterprise IT architectures are undergoing a historic transformation from “cloud-native” to “AI-native cloud” paradigms with the large-scale deployment of LLMs and AI agents. 🤖 We've been boasting expertise in Model Security and ModelOps lifecycle governance. 🔏 In the meantime, we've built systematic research capabilities for cloud AI component risks, systematically mapped the fingerprints and attack surfaces of 100+ AI components.

💡 There's no isolated or patchwork defense mechanism can effectively address the uncertain threats of the AI era. NSFOCUS will further strengthen the cooperation with ecosystem partners across the industry chain, jointly advancing the development of the global AI security industry.

06/03/2026

To accurately analyze the evolving landscape of global Advanced Persistent Threats ( ) and bolster defenses for digital security and critical information infrastructure, NSFOCUS has released the 𝟮𝟬𝟮𝟱 𝗔𝗣𝗧 𝗔𝗻𝗻𝘂𝗮𝗹 𝗟𝗮𝗻𝗱𝘀𝗰𝗮𝗽𝗲 𝗥𝗲𝗽𝗼𝗿𝘁, combining robust cybersecurity monitoring and advanced threat hunting capabilities.

📜 This report conducts an in-depth analysis of APT activities across 2025 and decodes cutting-edge attack techniques, delivering actionable guidance to defend against APT activities.

In 2025, we recorded 308 APT incidents throughout the year, representing a YoY increase of 4%. Global APT activities showed the characteristics of "technical sophistication, tactical complexity, and frequent vulnerabilities". 👿 Driven by new attack strategies and new productivity tools, APT groups’ attack techniques and tactics continued to upgrade, and their attack accuracy and destructive power have been significantly improved.

👉 Download the full report: https://nsfocusglobal.com/resources/2025-apt-annual-landscape-report/

05/29/2026

LLMs represented by 𝗠𝘆𝘁𝗵𝗼𝘀 have greatly reduced the cost and skill threshold for discovering and exploiting vulnerabilities, making advanced cyber attack capabilities that previously required national resources within reach.

⚠️ The structural shift is real: the time from vulnerability disclosure to weaponization is being compressed to near-zero. Defenders relying on traditional response timelines are already at a systemic disadvantage.

We've published a report analyzing the AI threat landscape following the next-gen LLMs capable of autonomous vulnerability discovery and exploitation.

What should enterprises do?
✅ Build AI + Blue Army platforms for normalized attack-defense drills
✅ Embed AI security agents directly into CI/CD pipelines
✅ Establish SBOM/AIBOM inventories for real-time 0-day impact assessment
✅ Shift focus from perimeter defense to limiting blast radius after breach
✅ Deploy AI-driven XDR platforms for machine-speed threat detection and response

The era of 𝘼𝙄 🆚 𝘼𝙄 𝙞𝙣 𝙘𝙮𝙗𝙚𝙧𝙨𝙚𝙘𝙪𝙧𝙞𝙩𝙮 is here. Read the full report to understand the threat landscape and our recommended response framework.

🔗https://nsfocusglobal.com/wp-content/uploads/2026/05/Security-Plan-for-AI-Threats.pdf

05/28/2026

In March 2026, NSFOCUS Security Lab discovered a total of 31 attack activities. These activities were mainly distributed in South Asia, Eastern Europe, and the Middle East.

3️⃣ The most active groups in March were from South Asia and based in Eastern Europe, while other relatively active groups included from the Middle East and from South Asia.
😈 87% of the total attack incidents used spear email attack as the intrusion method. A small number of threat actors also utilized vulnerability exploitations (10%) and watering hole attacks for infiltration (3%).
🪖 Military institutions were the primary targets in March, accounting for 33%, followed by government agencies (30%).
📌 In March, key incidents include orchestrating the Axios supply chain poisoning campaign, a vulnerability in products being escalated to a RCE severity, and leveraging OpenClaw-related GitHub repositories to distribute malware.

Learn more in 𝗡𝗦𝗙𝗢𝗖𝗨𝗦 𝗔𝗣𝗧 𝗺𝗼𝗻𝘁𝗵𝗹𝘆 𝗯𝗿𝗶𝗲𝗳𝗶𝗻𝗴:
🔗 https://nsfocusglobal.com/nsfocus-monthly-apt-insights-march-2026/

Regional APT Threat Situation In March 2026, the global threat hunting system of Fuying Lab detected a total of 31 APT attack activities. These activities were primarily concentrated in regions including South Asia, Eastern Europe, and the Middle East, as shown in the figure below. Regarding the act...

05/22/2026

In 2025, the ecosystem is experiencing accelerated fragmentation: while established threat actors continue to consolidate their dominance, emerging groups are rising swiftly by leveraging automation and intelligent capabilities, making the overall threat landscape increasingly complex.

Powered by long-term monitoring from NSFOCUS Fuying Lab’s Global Threat Hunting System, the newly released 𝟮𝟬𝟮𝟱 𝗚𝗹𝗼𝗯𝗮𝗹 𝗗𝗗𝗼𝗦 𝗔𝘁𝘁𝗮𝗰𝗸 𝗟𝗮𝗻𝗱𝘀𝗰𝗮𝗽𝗲 𝗥𝗲𝗽𝗼𝗿𝘁 provides a systematic breakdown of current DDoS trends.

📥 Download the full report to decode the trends and safeguard your infrastructure: https://nsfocusglobal.com/resources/2025-global-ddos-landscape-report/

05/20/2026

To overcome bottlenecks of AI models, including low performance and high memory usage in encrypted traffic detection, NSFOCUS and Intel jointly introduced 𝙁𝙖𝙨𝙩𝙟𝙤𝙮, a next-generation, open-source traffic feature analysis framework. 😉

Powered by Intel's hardware acceleration, VPP data-processing engine, and Intel® oneAPI Toolkit, is integrated into the NSFOCUS Unified Threat Sensor ( ) to deliver:

✅ Higher detection throughput at scale
✅ Real-time threat identification — without the memory overhead
✅ Lower device costs across gateway security deployments
✅ Extensible AI detection across NSFOCUS's broader product portfolio

We've just published a joint whitepaper detailing how this Intel-optimized pipeline — from hardware acceleration to AI model deployment — is redefining what's possible in encrypted traffic detection.

📖 Download a copy:https://nsfocusglobal.com/wp-content/uploads/2026/05/NSFOCUS-and-Intel-Achieve-High-Performance-Encrypted-Traffic-Detection-Through-Fastjoy.pdf

05/15/2026

🎉 Exciting news: NSFOCUS has been included in the Visionaries in the 2026 Gartner® Magic Quadrant™ for 𝗖𝘆𝗯𝗲𝗿𝘁𝗵𝗿𝗲𝗮𝘁 𝗜𝗻𝘁𝗲𝗹𝗹𝗶𝗴𝗲𝗻𝗰𝗲 𝗧𝗲𝗰𝗵𝗻𝗼𝗹𝗼𝗴𝗶𝗲𝘀!

For us, this is more than a milestone — it's a reflection of our commitment to turning threat intelligence into real, business-relevant action. 🧑‍💻

From AI-powered intelligence workflows, APT attribution research to our ThreatLens platform, we're focused on one thing: establish a full-spectrum threat intelligence product system spanning strategic, tactical, and technical levels. By deeply embedding threat intelligence into business contexts, we help enterprises advance their security operations from reactive response to proactive anticipation.

🔗 Read more:

SANTA CLARA, Calif., May 13, 2026 – On May 4, 2026, Gartner® published the Gartner® Magic Quadrant™ for Cyberthreat Intelligence Technologies report (hereinafter referred to as “the Report”). NSFOCUS was included in the Visionaries quadrant. We believe, this recognition reflects the intern...

05/13/2026

In 2025, fueled by the rapid evolution of AI agents and LLMs, attacks are undergoing a paradigm shift from traditional volumetric, bandwidth-heavy confrontations to intelligent warfare centered on cognitive speed, strategic precision, and decision-making efficiency.

📈 Attack methodologies have evolved from blunt traffic suppression to highly targeted precision strikes, resulting in a marked increase in both stealth and destructive impact.

🎯 The overall threat landscape is increasingly complex. As cyber and physical worlds become more deeply coupled, DDoS is evolving into a strategic geopolitical tool.

𝙄𝙨 𝙮𝙤𝙪𝙧 𝙙𝙚𝙛𝙚𝙣𝙨𝙚 𝙨𝙩𝙧𝙖𝙩𝙚𝙜𝙮 𝙧𝙚𝙖𝙙𝙮 𝙛𝙤𝙧 𝙩𝙝𝙚 𝘼𝙄 𝙚𝙧𝙖?

📥 Download the full report to decode the trends and safeguard your infrastructure:

THANK YOU FOR YOUR INTEREST IN NSFOCUS REPORTS​ 2025 Global DDoS Attack Landscape Report In 2025, fueled by the rapid evolution of AI agents and LLMs, DDoS attacks are undergoing a paradigm shift from traditional volumetric, bandwidth-heavy confrontations to intelligent warfare centered on cogniti...