Resources
Practical writing, research, and frameworks on AI security, governance, and getting more from what you spend.
NewsletterSeptember 9, 2026
A database connector says “read only.” Then a query it was supposed to stop gets through. A newly disclosed AWS vulnerability shows why the permissions behind an AI connector can matter more than the label in front of it.
Read the post
NewsletterAugust 14, 2026
A malicious agent skill does not necessarily need credentials of its own. A real 2026 campaign showed how seemingly legitimate instructions could route an authorized agent through secondary files and remote payloads until the workflow ended in credential theft, raising a harder question for enterprise security: what exactly must be trusted when instructions themselves can initiate the execution chain?
NewsletterJuly 17, 2026
An AI agent does not need to ignore its instructions to be compromised. New research shows how forged identities, validation fields, resource identifiers, and other trusted metadata can alter the environment an agent believes it is operating in, then steer legitimate tools toward unintended actions.
NewsletterJune 18, 2026
Most prompt-injection attacks are confined to a single interaction. Long-term memory changes that assumption. New research shows how information written under one user, task, or authorization context can survive for weeks, reappear later, and influence decisions long after the original interaction is gone.
NewsletterMay 15, 2026
AI agents may not expose their reasoning, but their tools leave a trail. New MCP security research is treating tool-call sessions as a detection surface, where the sequence of file reads, database queries, emails, errors, and arguments can reveal misuse that a single prompt log might miss. The catch is that the same traffic used for detection may also contain secrets, customer data, and execution paths. MCP tool-call monitoring is becoming useful evidence, but it also has to be handled like sensitive infrastructure telemetry.
NewsletterMay 4, 2026
Anthropic says its unreleased Mythos Preview model found and exploited high-severity vulnerabilities across every major operating system and browser, then chose to restrict access rather than release it. Independent researchers reproduced much of the discovery work using models costing a fraction of a cent per thousand tokens. The article examines what that split between cheap discovery and frontier exploitation means for enterprise patching programs that were built around a slower cycle.
NewsletterMarch 10, 2026
AI coding assistants have moved beyond autocomplete. Claude Code Security can scan full repositories, verify vulnerability findings, and propose patches directly in the pull request workflow. That puts it alongside CI servers and build pipelines as a component with its own credentials, configuration surfaces, and access to sensitive code. Security teams that have not yet accounted for it in their supply chain governance probably should.
White paperMarch 9, 2026
Large Language Models are transforming enterprise productivity. They're also creating a data security problem that existing tools weren't designed to solve.
White paperFebruary 20, 2026
When enterprises evaluate GenAI security solutions, they typically receive feature matrices: detection capabilities, supported data types, and compliance certifications. These comparisons create a false equivalence between solutions with fundamentally different architectures.
NewsletterFebruary 19, 2026
LLMjacking takes a familiar attack pattern — stolen cloud credentials — and points it at a new target: managed LLM inference. Recent incident writeups document a repeatable workflow, from stolen keys to quiet AI API probing to sustained model invocations that can drain budgets and exhaust quotas. For organizations where AI usage is growing faster than logging and cost controls, this attack class can turn a routine credential leak into an operational incident quickly.
NewsletterFebruary 11, 2026
BitBypass hides one sensitive word as a hyphen-separated bitstream, then uses system-prompt instructions to make the model decode and reinsert it. In testing across five frontier models, this approach substantially reduced refusal rates and bypassed multiple guard layers. All five tested models produced phishing content at rates between 68-92%. If your safety controls assume plain-language detection will catch malicious intent, this research deserves close attention.
NewsletterFebruary 11, 2026
A user clicks a Copilot link, watches it load, and closes the tab. The session keeps running. The data keeps flowing. Reprompt demonstrated what happens when AI assistants inherit user permissions, persist sessions silently, and cannot distinguish instructions from attacks. The vulnerability was patched. The architectural pattern that enabled it, ambient authority without session boundaries, still exists elsewhere..
NewsletterJanuary 23, 2026
A single poisoned document. An agent following instructions it should have ignored. An audit log that points to the wrong person. AI agents are no longer just automation: they're privileged identities that can be manipulated through their inputs. Regulators are catching up. NIST is collecting security input, FINRA is flagging autonomy and auditability as governance gaps, and Gartner predicts 25% of enterprise breaches will trace to agent abuse by 2028. The question isn't whether agents create risk. It's whether your controls were built for actors that can be turned by a document.
NewsletterJanuary 16, 2026
Enterprise AI breaches are not happening at the model layer. They are happening in the plumbing: context assembly, retrieval pipelines, tool orchestration. This article breaks down five documented failure modes, walks through a realistic attack scenario, and explains why prompt injection has become OWASP's top GenAI risk. Worth a read for anyone building or deploying AI systems with access to internal data.
NewsletterJanuary 12, 2026
Model confusion exposes an AI supply-chain risk hiding in plain sight. Code that appears to load a local model can silently resolve to a public registry model with the same name, opening the door to remote code execution or silent compromise. The risk lives in everyday ML code paths, not infrastructure, turning model loading itself into a security boundary enterprises rarely treat as one.
White paperOctober 23, 2025
See how G360 and Microsoft help organizations unlock AI innovation through practical, outcome-based modernization. Download the free e-book.
BlogAugust 25, 2025
Learn how G360 uses AI co-generation to help organizations transform legacy systems into cloud-native, scalable solutions built for the future.
BlogAugust 2, 2025
Discover how agentic AI systems bring autonomous decision-making and scale to Master Data Management, transforming data quality and stewardship.
BlogJuly 17, 2025
In 2025, AI integration and strict privacy regulations make data governance non-negotiable. Learn why governing your data is a strategic business imperative.
BlogJune 6, 2025
Explore how workflow automation tools eliminate bottlenecks, streamline approvals, and enable real-time collaboration to boost business productivity.
BlogMay 14, 2025
Learn how AI transforms Business Intelligence from static reporting to real-time, predictive, and prescriptive insights — and how to overcome integration challenges.
White paperMay 10, 2025
The future isn't coming, it's already here. But companies are still running on legacy systems built decades ago. These systems were never designed for today's speed, scale, or cloud infrastructure.
Start with a roadmap, a readiness assessment, or a structured takeover.