Security Assessment of DeepSeek Harness with A.I.G: Evaluating Resistance to Indirect Prompt Injection
We assess indirect prompt injection in DeepSeek Harness (DSH), using AI-Infra-Guard (A.I.G) to construct tests, deliver controlled taint, execute DSH, collect traces, and judge outcomes. The study covers 14,560 controlled executions over 16 indirect-content channels, text and file carrier modes, 35 payload objectives, one unmodified baseline, and 12 attack methods. The experiment preserves DSH's agent loop, tool registry, model adapter, and session-event path; source tools and sensitive sinks ar
Read on arXivLatest papers
- FinRCA-Bench: Benchmarking Evidence Retrieval and Reasoning for Financial AI SystemsAI often names the right cause of a financial mismatch without ever finding the proof for it
- Looped Language Models Improve Compositional Tool CallingAI models that rethink their own answers multiple times get better at chaining tools together
- FACET: Preserving Source Intent and Executable State in Terminal Task SynthesisFACET builds internally consistent terminal-task 'exam sets' to train command-line AI agents
- FM-Bench: A Benchmark for Long-Horizon Management with Competing AgentsHanding an AI a football club to run for 20 years reveals that winning comes from management habits, not raw model power
- Bridging Search and CRM: Productionizing AI Product Research Agents for Customer Re-EngagementAI agents win back window-shopping customers by chasing them down on WhatsApp
- Adversarial Review: Structured Disagreement for Grounded Agentic Code ReviewFor AI code review, one reviewer plus one critic beats piling on more agents
- GenEx: A Graph-Based Representational Paradigm for SARS-CoV-2 Variant Detection via Codon Co-occurrence NetworksTurning viral gene sequences into codon relationship maps to tell coronavirus variants apart
- Alignment Is All You Need: Instruction-Free Training for General Audio-Language ModelsA frozen language model plus one lightweight connector is enough to build a capable audio-understanding AI
Latest from METAL LAB
- NVIDIA's 300 Verified Skills Lift Correctness by 41 Points
- Wave your hand at a webcam, hear a theremin: browser instrument released
- Meta AI launches desktop app for Mac, can read an entire app window
- Factory Commits $100M to Partner Network, Pushes to Scale Software Factories
- SpaceX approached Cognition for acquisition four days after closing Cursor deal