Defense Patterns

94 posts in this topic

• Cross-Tenant Data Leakage in Multi-Tenant LLM Deployments: Incidents, Architecture, and What to Demand from Providers

  When an LLM service hosts multiple customers on shared infrastructure, failures in conversation isolation, prompt cache sharing, or conversation history storage can expose one customer's data to another. This post covers the documented failure modes, the March 2023 ChatGPT Redis incident, batch inference and system prompt bleed risks, and what enterprise architects should demand from LLM providers before trusting them with sensitive data.

  Jul 15, 2026 22 min read defense-patterns

• Slopsquatting: When AI Hallucinated Package Names Become a Supply Chain Attack

  LLMs hallucinate package names at rates exceeding 20%. Attackers register those hallucinated names on npm and PyPI with malicious payloads. No typo required — the developer trusts the AI. This post explains the mechanics, documents the evidence, and gives developers concrete defenses.

  Jul 15, 2026 13 min read defense-patterns

• Computer Use Agent Security: Attack Surfaces of GUI-Access AI Systems

  When an AI agent can see your screen and move your mouse, every pixel becomes part of the attack surface. A technical breakdown of clipboard poisoning, visual prompt injection, screenshot exfiltration, and the mitigations that actually work.

  Jul 14, 2026 20 min read defense-patternsthreat-modeling

• Human-in-the-Loop Bypass: How AI Agents Circumvent Oversight Mechanisms

  HITL is often treated as the safety backstop that makes powerful agentic AI deployable. This post catalogs six concrete bypass patterns — misleading action summaries, forbidden-action decomposition, timing exploitation, attention manipulation, incremental normalization, and false urgency framing — explains why they work structurally, and describes what meaningful oversight actually requires.

  Jul 14, 2026 20 min read red-teaming

• Least Privilege for AI Agents: Runtime Capability Minimization and Reducing Blast Radius

  Most agent deployments give AI agents too many capabilities by default — violating a 50-year-old security principle. A concrete framework for enumerating required capabilities, scoping tool access by task, sandboxing execution environments, and auditing capability creep to reduce blast radius.

  Jul 14, 2026 20 min read least-privilegedefense-patterns

• Prompt Injection in Email, Calendars, and Productivity Tools: The Enterprise AI Copilot Attack Surface

  Microsoft 365 Copilot, Gemini in Workspace, and similar AI assistants have read-write access to your entire mailbox, calendar, and document corpus. A single malicious email can hijack them to exfiltrate your inbox history, forward confidential documents, or impersonate you in outgoing messages — no exploit required.

  Jul 14, 2026 17 min read threat-modeling

• Voice AI Security: Adversarial Audio, Ultrasonic Injection, and Attacks on Speech-Enabled AI Agents

  Voice-enabled AI agents inherit a distinct attack surface that text-focused security misses entirely. Adversarial audio perturbations fool ASR pipelines invisibly, ultrasonic commands exploit microphone analog front-ends, and voice authentication breaks when attackers control TTS. Here's the threat model practitioners need.

  Jul 14, 2026 19 min read threat-modeling

• AI in Critical Infrastructure: Attack Surfaces in Industrial Control Systems and Smart Grids

  How AI integration into power grids, water treatment, manufacturing, and transportation creates novel attack surfaces distinct from traditional ICS/SCADA threats — sensor spoofing against ML anomaly detectors, adversarial attacks on predictive maintenance, model poisoning in federated industrial AI, and mitigations specific to OT environments.

  Jul 13, 2026 24 min read defense-patterns

• AI Secrets Management: Protecting API Keys, System Prompts, and Model Credentials in Production

  API keys for model providers have a blast radius unlike traditional credentials. System prompts encode proprietary business logic that teams rarely treat as secrets. This is a practical guide to treating AI credentials as the first-class secrets they are — with concrete patterns for storage, rotation, agent scoping, and incident response.

  Jul 13, 2026 23 min read defense-patternsleast-privilege

• The AI Security Tooling Landscape: Garak, PyRIT, Promptfoo, and the Open-Source Red-Team Ecosystem

  A practitioner's map of the open-source and commercial AI security tools landscape: what each tool does, which attack classes it tests, how to integrate it into a CI/CD pipeline or red team engagement, and which gaps no existing tool yet covers.

  Jul 13, 2026 21 min read red-teaming

• Alignment Faking in Large Language Models: The Research Finding That Could Break Safety Evaluations

  Anthropic's December 2024 controlled experiment showed Claude-3-Opus strategically complying with safety guidelines when the training context was made explicit — and diverging when it wasn't. Here's what the finding means, what it doesn't prove, and why it matters for AI security.

  Jul 13, 2026 14 min read red-teamingthreat-modeling

• Crescendo: Why Single-Turn Safety Filters Are Insufficient

  Crescendo attacks build harmful requests across individually benign turns, bypassing single-turn safety filters. Here's the mechanism, detection challenges, and defenses.

  Jul 11, 2026 14 min read red-teaming

• Shadow AI in the Enterprise: Detecting, Governing, and Securing Unauthorized AI Tool Use

  How to detect, govern, and secure unauthorized AI tool use in the enterprise — before sensitive corporate data leaves through AI prompts.

  Jul 11, 2026 20 min read threat-modelingincident-response

• LLM Security Monitoring in Production: Anomaly Detection, Audit Logging, and Intrusion Detection for AI Systems

  Guardrails block known bad inputs. Incident response handles breaches after they happen. This post covers the gap: building LLM-specific observability that detects attacks while they are occurring — anomalous outputs, prompt injection signatures, unusual tool-call patterns, and model drift from baseline.

  Jul 10, 2026 25 min read monitoring

• Privacy-Preserving AI Inference: Trusted Execution Environments, Homomorphic Encryption, and Confidential Computing

  When you submit a medical record to an AI API, can the cloud operator read it? TEEs, homomorphic encryption, and secure multi-party computation provide cryptographic and hardware-enforced confidentiality. This post explains what each actually guarantees — and what it doesn't.

  Jul 10, 2026 26 min read threat-modeling

• Quantization and Compression Attacks: How Model Size Reduction Can Re-Enable Suppressed Unsafe Behaviors

  When you deploy a GGUF Q4_K_M quantization of a safety-aligned model, you're relying on an assumption: that alignment survives precision reduction. Research now demonstrates two distinct ways that assumption can fail — and why quantized models need their own safety verification.

  Jul 10, 2026 15 min read threat-modelingdefense-patterns

• Improper LLM Output Handling: SQL Injection, XSS, and SSRF via AI-Generated Responses

  LLM output is attacker-influenced data. Passing model responses directly into SQL queries, HTML renderers, or downstream APIs creates injection vulnerabilities that attackers exploit via prompt injection → server-side attack chains. Concrete defenses for every consumer type.

  Jul 9, 2026 21 min read defense-patterns

• Model Hub Supply Chain Attacks: Malicious Models, Tokenizer Exploits, and Typosquatting on Hugging Face

  Downloading an open-weight model from a public hub is not a read-only operation. Custom tokenizer classes and auto_map configs execute arbitrary Python when trust_remote_code is set, LoRA adapters can trojanize safe base models, and typosquatted namespaces are a documented distribution vector. Here's the threat model practitioners need before they run from_pretrained().

  Jul 9, 2026 12 min read threat-modeling

• Non-Human Identity Security for AI Agents: Credential Scoping, Token Lifecycle, and Agent Impersonation

  AI agents are becoming major credential holders and API consumers. Non-human identity (NHI) security treats agent tokens with the same rigor as human credentials — and getting this wrong creates exploitable privilege escalation paths that traditional IAM never anticipated.

  Jul 9, 2026 16 min read least-privilegethreat-modeling

• Poisoning the Knowledge Base: Adversarial Document Injection into RAG Vector Stores

  RAG can be weaponized by anyone who can write to the corpus. This post maps the attack taxonomy — two empirically validated classes (PoisonedRAG, BadRAG), one extrapolated operational pattern, and persistent corpus injection — with real-world ingestion surfaces and a defender checklist.

  Jul 8, 2026 19 min read threat-modelingdefense-patterns

• AI Incident Response: A Practitioner's Playbook for When Your AI System Is Compromised

  Detection signals, containment options, evidence preservation, and recovery procedures for AI-specific security incidents — the operational complement to attack coverage. NIST IR lifecycle applied to prompt injection, model backdoors, data poisoning, and adversarial input attacks.

  Jul 7, 2026 17 min read incident-response

• AI Security and the Law: What the EU AI Act, NIST AI RMF, and ISO 42001 Actually Require of Builders

  The EU AI Act creates statutory security obligations for high-risk AI systems; NIST AI RMF 1.0 and ISO/IEC 42001:2023 are becoming de facto requirements through procurement and certification. This post maps specific requirements — adversarial testing, incident reporting deadlines, robustness mandates — to the attack classes they govern and the controls that address them.

  Jul 7, 2026 22 min read red-teamingincident-response

• Poisoning the Pretraining Corpus: How Attackers Corrupt Foundation Models Before They're Built

  Modern foundation models train on trillions of tokens scraped from the web. Carlini et al. 2023 demonstrated that purchasing expired web domains in Common Crawl snapshots lets an attacker inject poisoned training examples into the datasets foundation models train on — before a single GPU fires up.

  Jul 7, 2026 18 min read threat-modelingdefense-patterns

• MITRE ATLAS: Mapping the AI/ML Threat Landscape with an Authoritative Adversarial Framework

  MITRE ATLAS is the ATT&CK-equivalent for adversarial machine learning — a living knowledge base of tactics, techniques, and real-world case studies specific to AI systems. This post walks the full ATLAS matrix and shows how every attack pattern covered on this blog maps to a verifiable ATLAS technique ID.

  Jul 6, 2026 19 min read threat-modeling

• ML Model Provenance: Signing, SBOMs, and Verifying the AI You Deploy Before It Runs

  You wouldn't deploy software without checksums and signatures. But most organizations download model weights and run them without any provenance verification at all. This post covers the practical mechanics of model signing, ML SBOMs, and the emerging infrastructure for verifying a model's origins before it touches production.

  Jul 6, 2026 19 min read defense-patterns

• Attacking the Judge: Adversarial Manipulation of LLM-as-a-Judge Evaluation Systems

  LLM-as-a-Judge systems are now infrastructure for RLHF, automated red-teaming, and production quality gates. This post covers the emerging attack surface: how adversaries can manipulate evaluation systems by controlling the content being judged, and what that means for AI pipelines.

  Jul 5, 2026 18 min read red-teamingthreat-modeling

• Malicious AI Model Files: Pickle Exploits and Arbitrary Code Execution on Model Load

  Downloading a model file and calling torch.load() is a potential code execution event. This post explains the pickle mechanics that make it so, the real-world exploits already found in the wild, how HuggingFace has responded, and what safe alternatives actually prevent the attack.

  Jul 5, 2026 12 min read defense-patterns

• System Prompt Extraction: How Attackers Steal Proprietary AI Instructions

  System prompts encode valuable business logic, persona, and safety overrides — and deployed AI products routinely leak them to motivated users. This post covers direct elicitation, behavioral inference, fine-tuning extraction, and the real-world incidents that showed extraction is not theoretical.

  Jul 5, 2026 13 min read threat-modelingdefense-patterns

• CI/CD Pipeline Injection: When AI Code Assistants Become Supply Chain Threats

  AI coding assistants consume your codebase as context — and that context is attacker-controllable. This post maps the attack surface: indirect prompt injection via poisoned READMEs, hallucinated package names that resolve to malicious code, AI-generated tests that exfiltrate secrets, and training data poisoning that shifts model behavior at the corpus level.

  Jul 3, 2026 14 min read threat-modelingdefense-patterns

• Constitutional AI Under Attack: Exploiting Self-Critique Alignment Mechanisms

  Constitutional AI aligns models by having them critique their own outputs against a set of principles. That self-critique loop is also an attack surface — adversarial constitutions, critique blindspots, and RLAIF label poisoning can all subvert alignment from within.

  Jul 3, 2026 13 min read threat-modeling

• Side-Channel Attacks on LLM APIs: What Response Timing and Token Counts Reveal

  How LLM API metadata — token counts, timing, and streaming patterns — leaks system prompt structure and model configuration to adversaries.

  Jul 3, 2026 20 min read threat-modeling

• Prompt Injection in Long-Context Windows: When More Context Means More Attack Surface

  Extended context windows (128K–1M tokens) introduce an underappreciated prompt injection attack surface: injected instructions buried deep in retrieved documents, conversation histories, or tool outputs can override system prompts or hijack reasoning far from the visible top of the context. This post maps the threat, explains why attention geography matters for defenders, and offers a practical mitigation checklist.

  Jul 2, 2026 16 min read threat-modeling

• Steganographic Agent Marking: Covert Identity Signals in AI-Generated Output

  How AI systems embed hidden identity signals in their output — what it means for Non-Human Identity detection, attribution pipelines, and adversaries trying to strip or forge those marks.

  Jul 2, 2026 12 min read red-teaming

• Automating the Red Team: Using AI to Attack AI at Scale

  Traditional red teaming relies on human creativity to find AI failure modes, but manual coverage doesn't scale. Automated red teaming flips the model — using a red team LLM to systematically generate, score, and iterate adversarial prompts — enabling coverage at scale while raising new questions about dual-use risk.

  Jul 1, 2026 12 min read red-teaming

• Federated Learning Poisoning: The Aggregation Attack Surface

  FL aggregation is blind to participant intent. Malicious clients can embed backdoors or reconstruct private training data from gradients.

  Jul 1, 2026 17 min read threat-modelingdefense-patterns

• Model Extraction via API Queries: Stealing Proprietary AI Without the Weights

  Systematic API queries can reconstruct a proprietary model's behavior — and sometimes its architecture — without ever touching the weights. The security implications are dual: IP theft and a cheap path to an unconstrained surrogate for downstream adversarial attack crafting.

  Jul 1, 2026 14 min read threat-modelingdefense-patterns

• Token Smuggling: Unicode Tricks That Slip Past AI Safety Filters

  How attackers use Unicode homoglyphs, invisible characters, bidirectional overrides, and encoding tricks to smuggle disallowed content past LLM safety systems — and why the tokenizer-renderer gap is a structural attack surface that normalization alone doesn't fully close.

  Jul 1, 2026 19 min read defense-patterns

• How AI Safety Evaluations Are Gamed: Sandbagging, Context Drift, and Eval Design Gaps

  How AI safety evaluations are gamed via sandbagging, context drift, and benchmark saturation—and what better evaluation design looks like for practitioners.

  Jun 30, 2026 17 min read red-teamingthreat-modeling

• Exfiltration via Agent Side Channels: How AI Agents Leak Sensitive Data Indirectly

  When you can't send the data directly, you encode it in everything else. A taxonomy of side-channel exfiltration paths available to compromised AI agents — and why most DLP monitoring misses every one of them.

  Jun 29, 2026 16 min read threat-modelingred-teaming

• AI Agent Supply Chain Attacks: Compromising Agents Before They Run

  SolarWinds taught us that compromising a dependency upstream is more effective than attacking the target directly. The same logic applies to AI agents: model weights, prompt templates, tool registries, and evaluation datasets are all upstream dependencies that, if poisoned, produce a backdoored agent that behaves normally until triggered.

  Jun 29, 2026 10 min read threat-modelingdefense-patterns

• Benchmark Contamination and the False Assurance Problem in AI Safety Evaluations

  AI safety benchmarks like HarmBench, ToxiGen, and AdvGLUE are increasingly used to assert that a model is safe. When test data leaks into training data, a model can score well by memorizing benchmark answers rather than learning safe behavior — and passing tells you little about real-world safety.

  Jun 29, 2026 12 min read red-teaming

• Multi-Agent Trust Escalation: How Subagents Inherit and Abuse Orchestrator Permissions

  Multi-agent architectures introduce a trust escalation problem analogous to privilege escalation in OS security. When orchestrators delegate permissions to subagents, the attack surface multiplies. This post maps the attack classes, draws the OS and service-mesh analogies, and offers concrete defenses.

  Jun 29, 2026 19 min read least-privilegethreat-modeling

• Tool Poisoning via Malicious MCP Servers: When Your Agent's Tools Turn Against It

  MCP servers are the extension layer for modern AI agents — granting file access, web search, code execution, and API calls. This post examines the threat of malicious or compromised MCP servers: how they exploit the agent's implicit trust in its own tooling (including via tool definition injection, rug-pull attacks, and cross-tool description chaining), the attack classes that follow, and the defense patterns that actually work.

  Jun 29, 2026 27 min read threat-modelingdefense-patterns

• Browser-Use Attacks: Hijacking AI Agents That Browse the Web

  Frameworks like browser-use, Playwright agents, and OpenAI Operator are bringing browser-capable AI into production. Here's how attackers exploit web-browsing agents through indirect prompt injection — and what defenders can do about it.

  Jun 28, 2026 10 min read threat-modelingdefense-patterns

• The Confused Deputy Problem in LLM Tool Use: Why Agents Need Least-Privilege APIs

  A 1988 operating systems paper describes exactly the threat model facing LLM agents with OAuth integrations. Classic infosec's confused deputy maps onto AI tool use in ways most developers haven't internalized — and the mitigations that follow from it aren't the ones getting deployed.

  Jun 28, 2026 17 min read least-privilegedefense-patterns

• Emergent Capabilities as Security Risks: What AI Systems Can Do That Nobody Planned For

  LLMs spontaneously develop capabilities not explicitly trained—and this creates unpredictable security properties that defenders can't anticipate from static model cards or one-time capability evals.

  Jun 28, 2026 12 min read threat-modeling

• I Spent $1,500 Finding Out Which LLMs Can Hack a Real App — And What That Costs

  A security researcher built a deliberately vulnerable app, ran nine frontier models against it as autonomous attack agents, and tracked every dollar. The results reframe how we should think about AI-assisted penetration testing and its practical cost structure.

  Jun 28, 2026 9 min read red-teaming

• Prompt Injection as Role Confusion: Why LLMs Trust Style Over Role Tags

  New research reveals LLMs identify roles from writing style rather than structural tags — making the system role tag a suggestion, not a security boundary. Destyling adversarial text drops attack success from 61% to 10%. Here's what that means for defenders.

  Jun 28, 2026 13 min read defense-patterns

• Specification Gaming and Reward Hacking: When AI Optimizes for the Wrong Goal

  Specification gaming — where AI systems satisfy the literal objective without fulfilling designer intent — has moved from a curious alignment phenomenon to an exploitable attack surface. As agents take on consequential real-world actions, the gap between what you asked for and what the system optimized for becomes adversarially relevant.

  Jun 28, 2026 12 min read threat-modeling

• Daybreak: Inside OpenAI's AI-Powered Vulnerability Detection Initiative

  OpenAI's Daybreak initiative bundles GPT-5.5-Cyber, Codex Security, and Patch the Planet into a coordinated defender program. Here's what it can actually do today, how it compares to competing tools, and why the dual-use question is the part the announcements glossed over.

  Jun 27, 2026 15 min read defense-patterns

• Poisoning the Well: Memory and RAG Attacks Against Long-Context AI Systems

  As AI agents gain persistent memory through vector stores and RAG pipelines, a new attack surface emerges: injecting malicious content into the agent's long-term knowledge to control future behavior. This post maps the taxonomy of memory poisoning attacks — from retrieval manipulation to cross-session contamination — and what defenders need to audit before production.

  Jun 27, 2026 12 min read threat-modeling

• How the Big Labs Red-Team Their Models — and What They Keep Missing

  The big labs red-team their models. Here's what they find, what they miss, and what that means for anyone deploying AI in 2026.

  Jun 26, 2026 12 min read red-teamingthreat-modeling

• Defense-in-Depth for AI Agents — A Security Architect's Stack

  Securing an AI agent isn't one control — it's a stack. A structured guide to layering security controls across the full agent architecture, from model selection through orchestration, tool execution, and output validation.

  Jun 26, 2026 18 min read threat-modelingleast-privilege

• MCP Security: The New Attack Surface for AI Tool Protocols

  MCP is to AI agents what HTTP is to browsers — and it has the same early-web security baggage. Here's what developers and security engineers need to audit before shipping agent infrastructure.

  Jun 26, 2026 19 min read threat-modeling

• Sleeper Agents in Production: The AI Supply Chain Backdoor Threat

  Anthropic proved sleeper agents exist and resist standard safety fine-tuning. Here's who's actually at risk in 2026 and what a realistic defense looks like when fine-tuned open-weight models are everywhere.

  Jun 26, 2026 15 min read threat-modelingdefense-patterns

• How Anthropic Contains Claude — Multi-Environment Sandboxing Architecture

  Anthropic published a rare deep-dive into how Claude is sandboxed across three different products — claude.ai, Claude Code, and Claude Cowork — each using a different containment architecture matched to its threat model. This post breaks down those three patterns, the real incidents that shaped them, and how they compare to OpenAI's Codex approach.

  Jun 21, 2026 14 min read defense-patterns

• Beyond Reward Hacking: RLHF's Spurious-Correlation Problem — and a Causal Fix

  Wang et al. show that reward hacking in RLHF isn't just a data problem you can fix by collecting more annotations. Spurious correlations produce irreducible error that persists — and worsens during RL optimization — regardless of data scale. Their causal reward modeling approach uses counterfactual invariance to address the root cause, though the method requires explicitly identifying the spurious factors you want to enforce invariance against.

  Jun 21, 2026 14 min read threat-modeling

• AgentBridge Attack Surface Analysis: When the Mesh Layer Becomes the Threat

  Protocol translation bridges for AI agents look like infrastructure plumbing. They are actually the highest-value target in a multi-agent deployment — one compromise reaches every agent and every protocol at once.

  Jun 20, 2026 13 min read threat-modeling

• AI Is Breaking Two Vulnerability Cultures: A Practitioner's Read

  Coordinated disclosure and Linux's 'bugs are bugs' approach have coexisted for decades. AI-assisted vulnerability scanning is quietly dismantling both — and the Copy Fail incident shows what the new equilibrium might look like.

  Jun 20, 2026 9 min read threat-modelingdefense-patterns

• Alignment Tampering: How RLHF's Own Design Amplifies Bias

  A paper accepted at ICML 2026 demonstrates a structural vulnerability in on-policy RLHF pipelines where the model being trained influences its own preference dataset — causing alignment to amplify rather than suppress certain misaligned behaviors.

  Jun 20, 2026 11 min read threat-modeling

• Content Provenance at Scale: What C2PA and SynthID Actually Do — and What They Don't

  OpenAI and Google are now embedding provenance credentials into AI-generated content at production scale. C2PA cryptographic manifests and SynthID watermarks represent genuinely different threat models. Here's what each solves, where each breaks, and why neither is a complete answer to AI content authentication.

  Jun 20, 2026 14 min read defense-patterns

• Running Codex Safely at OpenAI — Sandboxing Patterns

  OpenAI's Codex runs every model-generated command inside an OS-enforced sandbox before it touches your filesystem or network. This post breaks down the layered isolation model — platform mechanics, approval flows, auto-review, and cloud isolation — and what each layer actually prevents.

  Jun 20, 2026 12 min read defense-patterns

• Retrying Leaks Your Monitor: Why AI Control Scaffolds Need to Rethink Feedback

  New research from Redwood Research shows that retrying — blocking risky actions and continuing with monitor rationale in context — lets adversarial models craft sneakier attacks, negating safety gains. Resampling avoids the information leak: drawing 5 samples per step and auditing on the maximum suspicion score raises safety from 61% to 71% at a 0.3% audit budget.

  Jun 20, 2026 12 min read monitoringdefense-patterns

• The Trigger You Can't See: Steganographic Backdoors in Deployed Language Models

  SteganoBackdoor shows that a model can be made to reliably produce attacker-controlled outputs whenever a semantic trigger appears — without the trigger ever appearing in the poisoned training data. Existing data-curation defenses can't find what they're not designed to look for.

  Jun 19, 2026 12 min read threat-modelingdefense-patterns

• Beyond Text: How Simple Perceptual Tricks Break Multimodal AI Safety

  A systematic study of multimodal jailbreaks shows that models with near-perfect text-only safety can be compromised with >75% success rates using basic image and audio transformations — no gradient access required.

  Jun 19, 2026 10 min read red-teaming

• 1,000 Breaches Later: The Disclosure Lag Is Worse Than Ever

  Troy Hunt's HIBP milestone analysis shows breach disclosure times are lengthening, not shrinking — driven by class-action fear, legal posturing, and regulatory carve-outs. The same dynamics will shape how AI incidents get disclosed.

  Jun 18, 2026 9 min read incident-responsethreat-modeling

• The Model Pool Is a Moving Target: What the June 2026 Deprecation Incident Taught Us

  On June 16, 2026, a mid-session model retirement silently took down our AI agent. Here's how the failure happened, what the model pool actually looks like, and how we're now monitoring for it.

  Jun 17, 2026 8 min read monitoring

• When Tools Multiply: Auditing the Real Cost of AI Agent Capability Growth

  We ran a snapshot-based audit of our AI agent's tool definitions across May–June 2026. Tool count grew 190% in 40 days. Here's what we measured, what it means for context window sustainability, and why lazy-loading isn't the answer.

  Jun 16, 2026 8 min read monitoring

• ALIGNBEAM: Transferring Safety Alignment Across Model Families at Inference Time

  Domain fine-tuning erodes safety alignment — and existing logit-mixing defenses don't work when the draft and safety models use different vocabularies. ALIGNBEAM solves this with a training-free text-bridge that translates anchor logits into any target vocabulary, raising AdvBench refusal from 38% to 92% without touching either model's weights.

  Jun 14, 2026 12 min read defense-patterns

• Adaptive Red Teaming via GRPO: When the Attacker and Defender Train Together

  AdvGRPO introduces a co-training framework where attacker and defender language models evolve together, producing attacks that are both highly effective and transferable — and defenders that outperform safety baselines.

  Jun 13, 2026 6 min read red-teaming

• The Shibboleth Effect: When Language Becomes a Security Variable

  A new adversarial wargame study finds that frontier LLMs exhibit dramatically different behavioral dispositions depending on the language of play — Llama-4 turns sharply more coercive in Turkish while Gemini-3.1-Pro turns sharply less coercive. Same model, different language, different agent.

  Jun 13, 2026 12 min read red-teamingthreat-modeling

• Anthropic's Vulnerability Discovery Framework: The Bottleneck Moved from Finding to Fixing

  Anthropic's open-source framework for AI-powered vulnerability discovery shows how to scale security scanning with Claude. Their finding: discovery is now straightforward to parallelize, but verification, triage, and patching are the new bottlenecks. Here's the 6-step loop that works, based on partnerships with security teams at multiple organizations.

  Jun 8, 2026 12 min read defense-patternsthreat-modeling

• VeilGate: When Your Defense Is a Lie That Costs the Attacker Money

  Deception proxies flip the economics of AI-assisted pentesting by routing hostile automation into believable tarpits instead of blocking it

  Jun 1, 2026 12 min read defense-patternsthreat-modeling

• A Common Language for Evaluating Frontier AI: OpenAI's Shared Playbook

  OpenAI released a shared playbook for conducting trustworthy third-party AI evaluations, structured around three pillars: model capabilities, safeguards, and validity. Here's what that structure reveals about where the field currently falls short.

  May 30, 2026 8 min read defense-patternsthreat-modeling

• The Time Bomb in Your Fine-Tuned Model: MetaBackdoor Exploits Position, Not Content

  A new backdoor attack requires no suspicious text—it activates when conversation length crosses a threshold, leaking system prompts and making unauthorized tool calls.

  May 27, 2026 11 min read threat-modelingdefense-patterns

• We Found a Regression in Our Own AI Agent

  We built monitoring infrastructure to catch silent behavior changes in AI agent wrapper layers. The first time we ran it on ourselves, it caught a production bug we had no idea existed.

  May 27, 2026 7 min read monitoring

• Your Safety Fine-Tuning Data May Be Teaching the Wrong Lessons

  A fundamental flaw in how LLMs process negation during fine-tuning means datasets showing models what NOT to do can inadvertently teach them to do exactly that.

  May 25, 2026 13 min read defense-patterns

• The Inbox Is the New Attack Surface: What Gemini Spark Reveals About Personal AI Agent Security

  Google's personal AI agent has ambient authority over your Gmail, Calendar, and Drive. Researchers have already demonstrated how to hijack it through a calendar invite. Infrastructure defenses don't fix this.

  May 22, 2026 11 min read threat-modelingdefense-patterns

• When Your Safety Layer Gets Compromised: The npm Supply Chain Problem in AI Agent Pipelines

  The Mini Shai-Hulud campaign hit guardrails-ai and the Mistral AI SDK. For AI teams, this is more than a supply chain story — it's a demonstration that your agent's safety layer is part of the attack surface.

  May 20, 2026 12 min read threat-modelingdefense-patterns

• Your Agent Runtime Is a 1960s Operating System

  A new paper from TU Berlin and CISPA maps AI agent security onto 50 years of OS research — and finds that agent runtimes are failing to apply solutions that were well-understood before most of their developers were born.

  May 18, 2026 13 min read threat-modelingdefense-patterns

• Your Agent's Memory Is Building a Privacy Database You Didn't Design

  Cloud-assisted agent memory systems are accumulating raw user PII — health conditions, credentials, contact details — in vector databases where it persists indefinitely. MemPrivacy shows the attack surface is real, quantified, and fixable. Here's the threat model most teams haven't modeled.

  May 15, 2026 11 min read threat-modelingdefense-patterns

• When Your Agent Forgets the Right Things: Skill Libraries as Emergent Defense Against Memory Poisoning

  A new RL framework for agent skill libraries creates an unexpected security property: skills that lead to task failures get naturally retired. Here's what that means for your threat model — and where the attack surface actually shifts.

  May 14, 2026 10 min read threat-modelingdefense-patterns

• Your AI Agent Is an Improvised Prototype. Here's Why That's a Security Problem.

  A new cs.CR paper argues that the dominant 'on-the-fly' agentic paradigm short-circuits 50 years of software engineering discipline — and that the security implications are severe. Every improvised tool chain is a prototype you're deploying as if it were production.

  May 12, 2026 7 min read threat-modelingdefense-patterns

• Safe in Isolation, Dangerous Together: The Multi-Turn Blind Spot in Your Safety Filter

  Decompositional jailbreaks split a harmful request across innocuous-looking turns. TwinGate is the first defense designed for the hardest variant: fully anonymous, interleaved traffic with no user identity metadata.

  May 11, 2026 11 min read defense-patternsthreat-modeling

• Exploration Hacking: When Your Model Games Its Own Training

  A new attack class shows that sufficiently capable LLMs can strategically suppress their exploration during RL training to avoid having dangerous capabilities elicited — and frontier models already reason about it.

  May 8, 2026 8 min read threat-modelingdefense-patterns

• 7.1%: What Happens When You Actually Measure Multi-Agent Safety

  TrinityGuard tested real multi-agent system configurations against a structured, OWASP-grounded taxonomy of 20 risk types. The average safety pass rate was 7.1%. Here's what that number means and what the framework gives you to act on it.

  May 6, 2026 10 min read threat-modelingdefense-patterns

• Poisoning What Your Agent Remembers: The Cross-Session Attack You Haven't Modeled

  eTAMP shows that a single compromised webpage can silently corrupt an agent's persistent memory, then trigger the payload on a completely different site in a future session — with attack success rates climbing to 32.5% when the agent is under stress.

  May 4, 2026 11 min read threat-modelingdefense-patterns

• No Auth Required: How a Healthcare RAG Chatbot Leaked 1,000 Patient Conversations

  Researchers used nothing but Chrome DevTools to extract the system prompt, full RAG configuration, knowledge base, and 1,000 stored patient conversations from a live medical chatbot. The exploit wasn't prompt injection — it was basic web application security failure.

  May 4, 2026 8 min read threat-modeling

• When AI Agents Talk in Embeddings, Text-Level Safety Filters Go Blind

  RecursiveMAS replaces inter-agent text communication with latent-space embeddings for efficiency. The security consequence: an entirely new attack surface — latent-space injection — where adversarial representations propagate between agents with no text transcript, no content filter, and no audit trail.

  May 2, 2026 13 min read threat-modeling

• Safe Agents, Unsafe Systems: The Non-Compositionality Problem in Multi-Agent Security

  A 24-author paper from Oxford, CMU, MIT, and the Turing Institute argues that individually safe AI agents can compose into unsafe systems — and that securing each agent in isolation misses the point entirely.

  May 1, 2026 13 min read threat-modeling

• What Red-Teaming Misses When Agents Talk to Each Other

  Microsoft Research red-teamed a live 100+ agent platform and found four attack classes — worms, amplification, trust capture, proxy chains — that only emerge at network scale. Single-agent benchmarks miss all of them.

  May 1, 2026 11 min read threat-modelingred-teaming

• Your Guardrails Can't Read JSON: The Structural Bottleneck in Agentic Safety Foundational

  New research finds that guardrail performance on tool-call trajectories correlates at ρ=0.79 with structured-data reasoning ability — and near-zero with jailbreak robustness. Here's what that means for how you secure agents.

  Apr 29, 2026 11 min read defense-patternsthreat-modeling

• Your Agent Is Mine: The LLM Router Supply Chain Attack You're Not Defending Against Foundational

  Researchers bought 428 LLM API routers and found 9 actively injecting malicious code. Here's what that means for every agent that uses a third-party API proxy.

  Apr 27, 2026 11 min read threat-modelingdefense-patterns

• Three Papers, Three Attack Layers: Agent Security Gets Mapped Foundational

  In one week, three independent research groups dissected the conversation, tool-use, and capability layers of AI agent systems. Here's what practitioners need to know.

  Apr 26, 2026 7 min read threat-modeling