Netcrook
AI Security & Agentic Systems / Asia
L’agente che ha dimenticato il cliente: perché l’AI enterprise si rompe ai margini
L’esperienza interna di AI di Zoho mostra che la vera debolezza dei sistemi agentici spesso non è il modello, ma il contesto aziendale, la validazione e le autorizzazioni che lo circondano.
When an AI Coder Becomes a Trust Problem, the Ban Comes Fast
A reported workplace block on Claude Code underscores a harder question for enterprises: who controls the assistant, what it can touch, and where its metadata travels.
Chinese LLMs Raise a New Question for Cyber Teams: Who Gets the Better Tools?
Two new models from Chinese firms are being discussed as serious rivals to top U.S. systems, and the security issue is not nationality alone but how much power those models get inside real workflows.
When an AI Coding Assistant Becomes a Governance Problem
A reported workplace ban on Claude Code shows how quickly agentic developer tools can turn from productivity aids into trust and auditability disputes.
When a Browser Prompt Becomes the Ransom Note
A new research claim points to an uncomfortable pattern: generative AI can help assemble ransomware-like workflows inside ordinary web pages, where one permission click may matter more than a downloaded file.
The Quiet Race for AI That Hunts Bugs Before Hackers Do
Chinese cyber-AI systems are narrowing the gap on vulnerability discovery, but the deeper shift is about who controls the models, the workflows, and the access around them.
When a Model Hunts Bugs, Policy Starts Chasing Code
Zhipu AI’s GLM-5.2 has been framed as a serious vulnerability detector, but the bigger story is how quickly cyber-capable AI is becoming a deployment problem, not just a research one.
When an Open-Weight Model Reaches Gated-Model Territory
A June release from Zhipu AI has put a new spotlight on AI-assisted vulnerability finding, where access controls, not just raw capability, are becoming the real policy fault line.
China’s New AI Bug Hunter Points to a Faster, Sharper Cyber Race
Tulongfeng is being framed as an AI system for finding software flaws and analyzing code, and that alone is enough to expose a growing bottleneck in modern defense: not discovery, but remediation.
When Search Starts Issuing Tasks: Naver’s AI Tab Pushes the Browser Into Action
Naver’s new AI Tab turns a familiar search box into a workflow layer, blending answers, maps, and reservation cues in a way that widens both convenience and the security surface.
When Workspace Boundaries Fail, AI Apps Start Leaking Sideways
Four flaws in Dify reportedly exposed weaknesses in tenant isolation, turning routine AI platform features into possible cross-workspace disclosure paths.
One Platform, Four Flaws: How DifyTap Turns AI Isolation Into a Weak Point
A cluster of Dify vulnerabilities shows how multi-tenant AI systems can leak across organizational boundaries when authorization checks fail at the control plane.
Japan’s Banks See a New Risk Layer: AI as a Force Multiplier for Cybercrime
The warning from Japan’s financial sector is not about a single breach, but about a harder problem: AI can make attacks faster, cheaper, and harder to contain.
SoftBank and OpenAI Turn Patch Speed Into a Cyber Product
A Japan launch tied to “Patching as a Service” suggests AI is moving from threat hunting into the slower, harder business of remediation.
Samsung’s Multi-Model Gamble Turns AI Governance Into the Real Security Story
The company is bringing ChatGPT, Gemini, and Claude into DX workflows, but the harder problem is not model choice - it is controlling data, permissions, and employee behavior.
VietBank’s Lean AI Bet Turns Banking Workflows Into a Security Decision
The bank is building internal AI for customer intelligence and office automation, but the real story is how data control, model choice, and cyber discipline now sit at the center of the design.
South Korea’s Cyber Agency Gets a Seat at Anthropic’s AI Security Table
KISA’s access to Mythos hints at a controlled, security-gated use of frontier AI where vulnerability hunting, not consumer chat, is the real prize.
Video AI’s New Power Move: Data, Distribution, and the Trust Problem
Chinese groups are gaining ground in generative video by pairing proprietary data with their own platforms, but the same formula can intensify deepfake, copyright, and provenance risks.
La vera battaglia dell’IA non è la capacità: è il controllo, il giudizio e il fallimento circoscritto
Un’intervista a un dirigente sulla strategia dell’IA punta a una lezione più profonda sulla sicurezza aziendale: la prossima ondata di IA si vincerà o si perderà nella governance, nella progettazione dei ruoli e nella capacità di innovare senza rompere i sistemi core.
Script Kiddies, Supercharged: How AI Turned Novices into Cybercrime Powerhouses
In 2026, artificial intelligence shattered the barriers to cyberattacks, unleashing a wave of sophisticated hacks by amateurs and upending digital defenses.