← American Express Interview Insights

American Express·AI Engineer·Technical Phone Screen·Senior

Senior
Apr 2026

Summary

Went through a technical screen for an AI Engineer role at American Express that was almost entirely focused on LLM security. Not what I expected from a fintech company but in hindsight it makes sense given how much they're investing in agentic systems. The depth they wanted was real.

Questions Asked (4)

Q1

What is the difference between direct prompt injection and indirect prompt injection, and why does the distinction matter for agent systems?

System DesignTechnical Trade-offs
Author's notes

I had a decent answer here.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clearly defining both types of prompt injection, emphasizing the source of the malicious input (direct from user vs. indirect from external data). Then explain why indirect injection is particularly dangerous in agent systems due to their autonomy and access to tools, and discuss mitigation strategies relevant to American Express's security-focused environment.

Pro tip: Tie the answer to real-world agent architectures like those using LangChain or AutoGPT, and mention how American Express's strict compliance and data protection requirements make indirect injection a critical risk to address.

1. Define Direct Prompt Injection

Explain that direct prompt injection occurs when a user deliberately inputs malicious prompts to manipulate the model's behavior, often to bypass restrictions or extract sensitive information.

2. Define Indirect Prompt Injection

Describe indirect prompt injection as when malicious content is embedded in external data sources (e.g., websites, documents, emails) that the agent processes, causing unintended actions without direct user input.

3. Highlight the Distinction

Emphasize that the key difference is the attack vector: direct is user-initiated, while indirect is data-initiated, making it harder to detect and more pervasive in agent systems that consume untrusted data.

4. Explain Why It Matters for Agent Systems

Discuss how agent systems autonomously interact with tools and external data, so indirect injection can lead to severe consequences like data exfiltration, unauthorized transactions, or system compromise, especially in high-stakes domains like finance.

5. Discuss Mitigation Strategies

Mention approaches like input sanitization, output validation, privilege separation, and monitoring for anomalous agent behavior, tailored to American Express's security and compliance needs.

Key Points to Mention

  • Direct injection relies on user input; indirect injection exploits external data sources.
  • Agent systems often have access to sensitive tools and data, amplifying the impact of indirect injection.
  • Indirect injection can be triggered without the user's knowledge, making it stealthier.
  • Real-world examples: malicious web pages causing agents to leak data or execute unauthorized actions.
  • Mitigation: sandboxing, strict input validation, and limiting agent permissions.
  • American Express context: financial regulations and data privacy make robust defenses essential.

AI-generated suggestions, not part of the candidate's original notes. May be inaccurate — verify before relying on them.

Q2

Walk through the main threat models you'd consider when deploying an LLM-based agent: data exfiltration, privilege escalation, jailbreaks, and tool misuse.

System DesignTechnical Trade-offsAdaptability & Ambiguity
Author's notes

Broad question and I tried to cover all four which was probably a mistake.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Structure your answer by first framing the agent's architecture and trust boundaries, then systematically address each threat model with concrete examples and mitigations. Emphasize defense-in-depth and how you'd balance security with usability in a regulated financial environment like American Express.

Pro tip: Tie each threat to a specific control (e.g., output sanitization for exfiltration, least-privilege tool access for escalation) and mention how you'd validate them with red-teaming and monitoring—this shows you think like an engineer, not just a theorist.

1. Map the agent architecture and trust boundaries

Briefly describe the LLM agent's components (model, tools, memory, external APIs) and where untrusted input enters. Identify trust boundaries to ground the threat discussion.

2. Analyze data exfiltration risks

Explain how sensitive data (PII, internal docs) could leak via model outputs, tool calls, or side channels. Discuss controls like output filtering, data minimization, and egress monitoring.

3. Examine privilege escalation vectors

Describe how an attacker could gain unauthorized access to tools or escalate permissions (e.g., prompt injection to invoke admin tools). Propose least-privilege design, scoped credentials, and human-in-the-loop for sensitive actions.

4. Address jailbreaks and prompt injection

Cover techniques like role-play, encoding tricks, and indirect injection via retrieved content. Suggest defenses: input sanitization, adversarial training, and output constraints.

5. Mitigate tool misuse and ensure safe operation

Discuss risks of tools being called with malicious parameters or unintended consequences. Recommend allow-lists, parameter validation, rate limiting, and audit logging.

Key Points to Mention

  • Defense-in-depth: layer controls across input, model, output, and tools
  • Least-privilege access for tools and data, with scoped API keys
  • Output sanitization and data loss prevention (DLP) for exfiltration
  • Prompt injection and jailbreak detection via input filtering and monitoring
  • Human-in-the-loop for high-risk actions (e.g., transactions, data access)
  • Continuous red-teaming and logging for audit and anomaly detection

AI-generated suggestions, not part of the candidate's original notes. May be inaccurate — verify before relying on them.

Q3

What mitigation strategies would you use to defend against prompt injection, and what are the concrete limits of each?

System DesignTechnical Trade-offs
Author's notes

This is where it got interesting.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Structure your answer as a layered defense-in-depth strategy, covering input validation, model-level safeguards, and output monitoring. For each mitigation, explicitly state its concrete limits—such as false positives, bypass potential, or performance overhead—to demonstrate technical maturity and risk awareness.

Pro tip: Emphasize that prompt injection is an unsolved problem and that the goal is risk reduction, not elimination. Mention that at a financial institution like American Express, you would also align mitigations with compliance and fraud prevention frameworks.

1. Define the threat model

Briefly classify prompt injection types (direct, indirect, jailbreak) and their potential impact on the system, such as data leakage or unauthorized actions.

2. Input sanitization and validation

Describe techniques like input filtering, encoding, and length limits to block obvious malicious patterns, and note their limits (e.g., bypass via obfuscation, false positives).

3. Model-level defenses

Explain approaches like prompt engineering, instruction hierarchy, and fine-tuning to make the model more robust, while acknowledging that no model is fully immune.

4. Output monitoring and guardrails

Discuss post-processing checks, anomaly detection, and human-in-the-loop for high-risk actions, and their limits (e.g., latency, scalability, evasion).

5. Continuous evaluation and adaptation

Highlight the need for red-teaming, monitoring, and updating defenses, noting that limits evolve as attackers innovate.

Key Points to Mention

  • Defense-in-depth: combine multiple layers (input, model, output) to reduce single-point failure.
  • Concrete limits: false positives, bypass via adversarial examples, performance overhead, and scalability challenges.
  • Prompt injection is an open problem; no mitigation is 100% effective.
  • Context matters: for a financial institution, prioritize data protection, compliance, and fraud prevention.
  • Trade-offs: balance security with usability, latency, and cost.
  • Continuous monitoring and red-teaming are essential to adapt to new attack vectors.

AI-generated suggestions, not part of the candidate's original notes. May be inaccurate — verify before relying on them.

Q4

How would you design an end-to-end defense strategy for an LLM agent that has access to external tools and retrieval systems?

System DesignTechnical Trade-offsAPI & Integrations
Author's notes

Saved the hardest for last.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Structure your answer around a layered defense-in-depth strategy covering input, agent reasoning, tool/retrieval, and output stages, with continuous monitoring and feedback loops. Emphasize how each layer mitigates specific risks like prompt injection, data leakage, and tool misuse, and tie your choices to American Express's regulatory and security requirements.

Pro tip: Frame your answer around business risk and compliance—mention specific regulations like PCI DSS and GDPR, and propose measurable security metrics (e.g., injection detection rate, false positive rate) to show you think like an engineer who understands enterprise constraints.

1. Threat Modeling & Risk Assessment

Identify potential threats across the agent's lifecycle: prompt injection, data exfiltration, tool abuse, retrieval poisoning, and model inversion. Prioritize risks based on impact and likelihood, considering American Express's sensitive financial data.

2. Input & Output Guardrails

Implement input sanitization, prompt injection detection (e.g., using classifiers or rule-based filters), and output filtering to prevent leakage of PII or sensitive information. Use techniques like context-aware redaction and allow-listing for tool calls.

3. Secure Tool & Retrieval Integration

Apply least-privilege access to tools, validate and sanitize retrieved documents, and use provenance tracking to detect poisoned data. For APIs, enforce authentication, rate limiting, and schema validation.

4. Agent Reasoning & Action Monitoring

Add runtime monitoring for anomalous agent behavior, such as unexpected tool sequences or excessive retrieval. Implement human-in-the-loop for high-risk actions and use canary tokens to detect data exfiltration attempts.

5. Continuous Evaluation & Incident Response

Set up logging, auditing, and automated red-teaming to continuously test defenses. Define incident response playbooks for breaches and feed learnings back into the system for iterative improvement.

Key Points to Mention

  • Defense-in-depth with multiple layers (input, reasoning, tool, output) to avoid single point of failure.
  • Prompt injection and jailbreak mitigation techniques, such as input sanitization and instruction hierarchy.
  • Least-privilege access and sandboxing for external tools and APIs.
  • Retrieval security: document validation, provenance, and poisoning detection.
  • Monitoring and observability: logging, anomaly detection, and canary tokens.
  • Compliance with financial regulations (PCI DSS, GDPR) and alignment with American Express's security policies.

AI-generated suggestions, not part of the candidate's original notes. May be inaccurate — verify before relying on them.