Roughly 1% of real resumes contain hidden prompt injections against LLM screeners, prevalence has risen over 1-2 years, and over 90% avoid explicit instructions.
hub Canonical reference
Securing AI Agents with Information-Flow Control
Canonical reference. 91% of citing Pith papers cite this work as background.
51
Pith papers citing it
2
external citations · Pith
Background
91% of classified citations
abstract
As AI agents become increasingly autonomous and capable, ensuring their security against vulnerabilities such as prompt injection becomes critical. This paper explores the use of information-flow control (IFC) to provide security guarantees for AI agents. We present a formal model to reason about the security and expressiveness of agent planners. Using this model, we characterize the class of properties enforceable by dynamic taint-tracking and construct a taxonomy of tasks to evaluate security and utility trade-offs of planner designs. Informed by this exploration, we present Fides, a planner that tracks confidentiality and integrity labels, deterministically enforces security policies, and introduces novel primitives for selectively hiding information. Its evaluation in AgentDojo demonstrates that this approach enables us to complete a broad range of tasks with security guarantees. A tutorial to walk readers through the the concepts introduced in the paper can be found at https://github.com/microsoft/fides
hub tools
citation-role summary
background 10
dataset 1
citation-polarity summary
representative citing papers
Oracle Poisoning corrupts knowledge graphs used by AI agents via tool calls, leading tested models to accept fabricated claims at 100% under directed queries in a production-scale demonstration.
MCPHunt benchmark finds 11.5-41.3% policy-violating credential propagation in multi-server MCP agents across five models, reducible up to 97% by prompt mitigations while retaining most utility.
NeuroTaint is the first taint tracking framework for LLM agents that uses offline auditing of semantic, causal, and persistent context to detect flows from untrusted sources to privileged sinks.
TRUSTDESC prevents tool poisoning in LLM applications by automatically generating accurate tool descriptions from code via a three-stage pipeline of reachability analysis, description synthesis, and dynamic verification.
Prismata cuts web-agent prompt-injection attack success from 85.5% to 0.7% via Biba-inspired DOM trust labeling and mechanical least-privilege confinement without site annotations.
Payload-preserving SkillCloak evasions defeat install-time skill scanners at ≥90% with SFS packing, while runtime SkillDetonate detects ~97% of attacks via sandbox taint evidence.
Presents TMA-NM, a non-malleable origin-bound authority system for LLM-agent memory with TLA+ machine-checked separation theorems and benchmarks showing 0% attack success against direct and laundering poisoning while preserving utility.
AutoDojo adaptively optimizes IPI attacks to bypass defenses, recovering substantial ASR on action-open tasks where static attacks fail.
The paper introduces Consent Integrity as the property that actions shown for approval must be rendered by a trusted mediator from the real boundary action over an unspoofable path and bound to execution, with uninspectable actions surfaced rather than silently approved.
Boiling the Frog is a new stateful multi-turn benchmark that finds an aggregate 44.4% strict attack success rate for incremental safety violations across nine AI models, with rates ranging from 20.5% to 92.9%.
PACT achieves perfect security and utility under oracle provenance by enforcing argument-level trust contracts based on semantic roles and cross-step provenance tracking, outperforming invocation-level monitors in AgentDojo evaluations.
ARGUS cuts context-aware prompt-injection success from 28.8% to 3.8% on AgentLure while keeping 87.5% clean utility, beating prior defenses on the security-utility tradeoff.
A parameterized DFA firewall enforces safe tool sequences for structured AI agents, reducing attack success rates to 2.2% in tested workflows with low added latency.
The paper defines causality laundering as an attack leaking information from denial outcomes in LLM tool calls and proposes the Agentic Reference Monitor to block it using denial-aware provenance graphs.
FORGE enforces security policies in agentic systems via Datalog over abstract predicates with an observability service and reference monitor that guarantees policy semantics when the environment contract holds.
CXI enforces that LLM agent tool calls execute only when protected sink fields, sink-interpreted effects, and the invocation event each carry authority bound to the same canonical action manifest.
MemLeak benchmark shows retained images enable 12% recovery of deleted facts in multimodal agents (reduced to 2% with content-aware deletion), with 47% of image leaks not text-recoverable.
ActPlane introduces an OS-kernel policy engine using an information-flow control DSL and eBPF to enforce agent harness policies, achieving better compliance on indirect paths with 1.9-8.4% overhead.
AOHP is an Android-based OS harness that reports 21% higher agent task completion, 52% lower token cost, and improved security compliance through agent-first system primitives.
GIF introduces a Jacobian-based upper bound on input-output mutual information in LLMs with formal Lean proof and strong empirical recall on injection and leakage benchmarks.
Activation probes, calibrated honeytokens, and multi-turn leakage accounting detect credential exfiltration attempts in LLM agents with high accuracy in controlled open-model tests.
Ghost tool calls from speculative dispatch create persistent intent leaks that only issue-time policies changing or suppressing call arguments or destinations can reduce, per evaluations of twelve policies on three corpora.
AuthGraph aligns an execution provenance graph with a clean authorization graph to detect parameter-source deviations from user intent, reducing attack success rates to 1-2% on AgentDojo and AgentDyn while retaining most task utility.
citing papers explorer
-
Measuring Real-World Prompt Injection Attacks in LLM-based Resume Screening
Roughly 1% of real resumes contain hidden prompt injections against LLM screeners, prevalence has risen over 1-2 years, and over 90% avoid explicit instructions.
-
Oracle Poisoning: Corrupting Knowledge Graphs to Weaponise AI Agent Reasoning
Oracle Poisoning corrupts knowledge graphs used by AI agents via tool calls, leading tested models to accept fabricated claims at 100% under directed queries in a production-scale demonstration.
-
MCPHunt: An Evaluation Framework for Cross-Boundary Data Propagation in Multi-Server MCP Agents
MCPHunt benchmark finds 11.5-41.3% policy-violating credential propagation in multi-server MCP agents across five models, reducible up to 97% by prompt mitigations while retaining most utility.
-
Ghost in the Agent: Redefining Information Flow Tracking for LLM Agents
NeuroTaint is the first taint tracking framework for LLM agents that uses offline auditing of semantic, causal, and persistent context to detect flows from untrusted sources to privileged sinks.
-
TRUSTDESC: Preventing Tool Poisoning in LLM Applications via Trusted Description Generation
TRUSTDESC prevents tool poisoning in LLM applications by automatically generating accurate tool descriptions from code via a three-stage pipeline of reachability analysis, description synthesis, and dynamic verification.
-
Prismata: Confining Cross-Site Prompt Injection in Web Agents
Prismata cuts web-agent prompt-injection attack success from 85.5% to 0.7% via Biba-inspired DOM trust labeling and mechanical least-privilege confinement without site annotations.
-
Cloak and Detonate: Scanner Evasion and Dynamic Detection of Agent Skill Malware
Payload-preserving SkillCloak evasions defeat install-time skill scanners at ≥90% with SFS packing, while runtime SkillDetonate detects ~97% of attacks via sandbox taint evidence.
-
Securing LLM-Agent Long-Term Memory Against Poisoning: Non-Malleable, Origin-Bound Authority with Machine-Checked Guarantees
Presents TMA-NM, a non-malleable origin-bound authority system for LLM-agent memory with TLA+ machine-checked separation theorems and benchmarks showing 0% attack success against direct and laundering poisoning while preserving utility.
-
AutoDojo: Adaptive Black-Box Attacks Reveal the Limits of IPI Defenses and Task-Specification Effects in LLM Agents
AutoDojo adaptively optimizes IPI attacks to bypass defenses, recovering substantial ASR on action-open tasks where static attacks fail.
-
What You Approve Is What Executes: Consent Integrity for Black-Box LLM Agents
The paper introduces Consent Integrity as the property that actions shown for approval must be rendered by a trusted mediator from the real boundary action over an unspoofable path and bound to execution, with uninspectable actions surfaced rather than silently approved.
-
Boiling the Frog: A Multi-Turn Benchmark for Agentic Safety
Boiling the Frog is a new stateful multi-turn benchmark that finds an aggregate 44.4% strict attack success rate for incremental safety violations across nine AI models, with rates ranging from 20.5% to 92.9%.
-
The Granularity Mismatch in Agent Security: Argument-Level Provenance Solves Enforcement and Isolates the LLM Reasoning Bottleneck
PACT achieves perfect security and utility under oracle provenance by enforcing argument-level trust contracts based on semantic roles and cross-step provenance tracking, outperforming invocation-level monitors in AgentDojo evaluations.
-
ARGUS: Defending LLM Agents Against Context-Aware Prompt Injection
ARGUS cuts context-aware prompt-injection success from 28.8% to 3.8% on AgentLure while keeping 87.5% clean utility, beating prior defenses on the security-utility tradeoff.
-
Enforcing Benign Trajectories: A Behavioral Firewall for Structured-Workflow AI Agents
A parameterized DFA firewall enforces safe tool sequences for structured AI agents, reducing attack success rates to 2.2% in tested workflows with low added latency.
-
Causality Laundering: Denial-Feedback Leakage in Tool-Calling LLM Agents
The paper defines causality laundering as an attack leaking information from denial outcomes in LLM tool calls and proposes the Agentic Reference Monitor to block it using denial-aware provenance graphs.
-
Formal Policy Enforcement for Real-World Agentic Systems
FORGE enforces security policies in agentic systems via Datalog over abstract predicates with an observability service and reference monitor that guarantees policy semantics when the environment contract holds.
-
Context-to-Execution Integrity for LLM Agents
CXI enforces that LLM agent tool calls execute only when protected sink fields, sink-interpreted effects, and the invocation event each carry authority bound to the same canonical action manifest.
-
MemLeak: Diagnosing Information Leaks in Multimodal Agent Memory
MemLeak benchmark shows retained images enable 12% recovery of deleted facts in multimodal agents (reduced to 2% with content-aware deletion), with 47% of image leaks not text-recoverable.
-
ActPlane: Programmable OS-Level Policy Enforcement for Agent Harnesses
ActPlane introduces an OS-kernel policy engine using an information-flow control DSL and eBPF to enforce agent harness policies, achieving better compliance on indirect paths with 1.9-8.4% overhead.
-
AOHP: An Open-Source OS-Level Agent Harness for Personalized, Efficient and Secure Interaction
AOHP is an Android-based OS harness that reports 21% higher agent task completion, 52% lower token cost, and improved security compliance through agent-first system primitives.
-
GIF: Locally Sound Geometric Information Flow Control for LLMs
GIF introduces a Jacobian-based upper bound on input-output mutual information in LLMs with formal Lean proof and strong empirical recall on injection and leakage benchmarks.
-
Caught in the Act(ivation): Toward Pre-Output and Multi-Turn Detection of Credential Exfiltration by LLM Agents
Activation probes, calibrated honeytokens, and multi-turn leakage accounting detect credential exfiltration attempts in LLM agents with high accuracy in controlled open-model tests.
-
Ghost Tool Calls: Issue-Time Privacy for Speculative Agent Tools
Ghost tool calls from speculative dispatch create persistent intent leaks that only issue-time policies changing or suppressing call arguments or destinations can reduce, per evaluations of twelve policies on three corpora.
-
Aligning Provenance with Authorization: A Dual-Graph Defense for LLM Agents
AuthGraph aligns an execution provenance graph with a clean authorization graph to detect parameter-source deviations from user intent, reducing attack success rates to 1-2% on AgentDojo and AgentDyn while retaining most task utility.
-
Securing LLM Agents Need Intent-to-Execution Integrity
The paper defines intent-to-execution integrity as the conjunction of Tool Integrity, Instruction Integrity, Judgment Integrity, and Data Flow Integrity, arguing that existing LLM agent defenses provide only partial coverage of these properties.
-
MemLineage: Lineage-Guided Enforcement for LLM Agent Memory
MemLineage enforces untrusted-path persistence in LLM agent memory through Merkle logs, per-principal signatures, and max-of-strong-edges lineage propagation, achieving zero ASR on three poisoning workloads with sub-millisecond overhead.
-
Web Agents Should Adopt the Plan-Then-Execute Paradigm
Web agents should default to planning a complete task program before observing live web content to reduce prompt injection exposure, since WebArena tasks are compatible and 80% need no runtime LLM calls.
-
Sleeper Channels and Provenance Gates: Persistent Prompt Injection in Always-on Autonomous AI Agents
Sleeper channels enable persistent prompt injection in always-on AI agents via persistence substrate and firing separation, countered by provenance gates using action digests and owner attestations with a soundness theorem.
-
Language-Based Agent Control
LBAC is a new programming model that enforces user-specified policies on agentic applications by requiring agent-generated programs to be well-typed in the context of the scaffolding code.
-
AgentShield: Deception-based Compromise Detection for Tool-using LLM Agents
AgentShield uses layered deception traps in LLM agent tool interfaces to detect indirect prompt injection compromises with 90.7-100% success on commercial models, zero false positives, and cross-lingual transfer without retraining.
-
Trojan Hippo: Weaponizing Agent Memory for Data Exfiltration
The paper defines and evaluates Trojan Hippo attacks on LLM agent memory, showing 85-100% success in data exfiltration across backends and reduced rates with defenses at varying utility costs.
-
Semia: Auditing Agent Skills via Constraint-Guided Representation Synthesis
Semia synthesizes Datalog representations of agent skills via constraint-guided loops to enable reachability queries for semantic risks, finding critical issues in over half of 13,728 real skills with 97.7% recall on expert-labeled samples.
-
Alignment Contracts for Agentic Security Systems
Alignment contracts define scope, allowed effects, budgets and disclosure rules as safety properties over finite effect traces, with decidable admissibility, refinement rules, and Lean-verified soundness under an observability assumption.
-
An AI Agent Execution Environment to Safeguard User Data
GAAP guarantees confidentiality of private user data for AI agents by enforcing user-specified permissions deterministically through persistent information flow tracking, without trusting the agent or requiring attack-free models.
-
Beyond Pattern Matching: Seven Cross-Domain Techniques for Prompt Injection Detection
Seven cross-domain prompt-injection detectors are introduced; three are shipped and d028 raises F1 on paraphrased attacks from 0.033 to 0.378, while adaptive-attack support remains unevaluated.
-
Don't Make Models Guess Security and Safety: Symbolic Guardrails for Domain-Specific AI Agents
Symbolic guardrails enforce about 74% of agent security and safety requirements on three benchmarks with mostly simple checks, improving safety without sacrificing utility.
-
Agents That Know Too Much: A Data-Centric Survey of Privacy in LLM Agents
A data-centric survey finds that only information-flow control covers compositional and cross-session leakage in LLM agents and that no single benchmark tests an agent across all its data surfaces under one policy.
-
VIGIL: Runtime Enforcement of Behavioral Specifications in AI Agent Skills
VIGIL introduces a policy language and symbolic evaluation rules to enforce context-aware behavioral specifications on LLM agent traces, achieving over 95% recall and under 10% false positives on real tasks.
-
Adaptive Evaluation of Out-of-Band Defenses Against Prompt Injection in LLM Agents
An independent reproduction on AgentDojo with Qwen2.5-7B finds that the Progent out-of-band defense reduces mean attack success from 25.8% to 4.2% and holds against a hand-crafted adaptive attack at 2.6%.
-
Governed Shared Memory for Multi-Agent LLM Systems
The paper defines four governance primitives for shared memory in multi-agent LLM fleets and reports evaluation results from a live production service.
-
From Agent Traces to Trust: A Survey of Evidence Tracing and Execution Provenance in LLM Agents
This survey defines execution provenance as a typed graph of agent execution and evidence tracing as its projection onto evidence-support relations, then reviews methods, taxonomy, benchmarks, and challenges for auditable LLM agents.
-
ChainCaps: Composition-Safe Tool-Using Agents via Monotonic Capability Attenuation
ChainCaps uses monotonic capability attenuation via intersection of sink-specific budgets in a transparent proxy to reduce attack success on composed tool-using agents from 25-68% to 0-4.8% while keeping 96-100% benign task completion.
-
Reframing LLM Agent Security as an Agent-Human Interaction Problem
LLM agent security is reframed as an agent-human interaction issue, supported by a survey showing industry preference for human-centric mechanisms over academic favorites and proposing a new research agenda.
-
ADR: An Agentic Detection System for Enterprise Agentic AI Security
ADR is a three-component detection system for AI agents that combines telemetry sensors, red teaming, and two-tier detection, achieving 97.2% precision in a ten-month Uber deployment and outperforming baselines on the new ADR-Bench.
-
The Misattribution Gap: When Memory Poisoning Looks Like Model Failure in Agentic AI Systems
Memory poisoning via lost-provenance documents in agent memory stores creates agent misconduct that safety systems misattribute to model failure; the paper defines Semantic Norm Drift, releases a benchmark, and proposes a new testing method plus a defense.
-
Can LLMs Make (Personalized) Access Control Decisions?
LLMs reflect users' privacy preferences in access control decisions with up to 86% agreement and can promote safer behavior, but personalization trades off higher individual match for potentially less secure results when users over-permission.
-
Assessing Automated Prompt Injection Attacks in Agentic Environments
Black-box optimization outperforms gradient-based methods for prompt injection on LLM agents, with success depending on attacker model strength and limited transfer from small to frontier models.
-
Agent Security is a Systems Problem
The paper argues that agent security is best addressed as a systems problem by applying principles from operating systems, networks, and formal methods rather than relying solely on model robustness improvements.
-
Engineering Robustness into Personal Agents with the AI Workflow Store
Position paper advocating a shift from on-the-fly AI agent synthesis to reusable hardened workflows in an AI Workflow Store to improve robustness and security.
-
Toward Secure LLM Agents: Threat Surfaces, Attacks, Defenses, and Evaluation
A synthesis of 247 papers on LLM agent security identifies prompt injection and tool hijacking as dominant threats, notes weakly compositional defenses, and argues for trust boundaries and realistic evaluations.