Advisor, AI Ethics

What matters most

• Independent AI Ethics reviews

  The role conducts and documents reviews of new and existing AI projects, using evaluation, testing, red-teaming and assurance methods to assess potential harms. Prepare to explain how a review would turn evidence into documented findings.

• Consumer-product risk assessment

  The primary focus is consumer-facing AI products. Reviews cover unfair treatment, transparency, safety, misuse and human oversight; generative and agentic AI risks are preferred areas of knowledge. Prepare a structured approach to assessing those risks in real-world use.

• Lifecycle advice across teams

  The advisor provides sustained support to an assigned product portfolio and works with business and technical teams to align development, deployment and monitoring with Lenovo’s AI Policy and Responsible AI principles. Prepare to discuss how advice changes as a product moves through its lifecycle.

• Improving and sharing governance practices

  The role improves review criteria, tools and documentation; partners with the Inclusive Product Design Office; and develops internal Responsible AI training. Prepare to distinguish a project-specific review from a reusable framework or training improvement.

Likely challenges to think through

• Balancing product utility and potential harm

  Likely challenge, not a stated employer problem: consumer-facing AI may offer useful features while creating safety, misuse or oversight risks. Prepare to describe how you would assess severity and likelihood, propose mitigations, and document residual risk without assuming every risk has a simple pass-or-fail answer.

• Maintaining independent judgment while advising teams

  Likely challenge: the role calls for independent reviews and sustained cross-functional support. The trade-off is giving candid findings while remaining useful to teams making product decisions. Prepare an example approach for communicating evidence, uncertainty and actionable recommendations.

• Keeping reviews consistent as risks evolve

  Likely challenge: global regulations, standards, evaluation methods and AI risks change, while review processes need usable criteria. Prepare to explain how you would update criteria and documentation without making reviews inconsistent or overly burdensome.

Topics to prepare

• AI governance across the product lifecycle

  Posting: The advisor helps align development, deployment and monitoring with Lenovo’s AI Policy, ethical principles and regulatory requirements. Exercise: Outline what an ethics review should consider at each lifecycle stage; do not assume the contents of Lenovo’s internal policy.

• Sociotechnical risks in consumer-facing AI

  Posting: Reviews consider unfair treatment, transparency, safety, misuse and human oversight; systems thinking is preferred. Exercise: For a hypothetical consumer AI feature, map users, affected people and real-world contexts, then explain how each could change the risk assessment.

• Generative and agentic AI risks

  Posting: Understanding these capabilities and their risks, particularly in consumer products, is preferred. Exercise: Compare potential failure and misuse scenarios for a generative feature and an agentic feature, including where human oversight matters.

• AI assurance methods

  Posting: Reviews use evaluation, assessment, testing, red-teaming and assurance methods. Exercise: Explain what evidence each method can provide for a stated harm and what uncertainty would remain.

• Regulatory and standards awareness

  Posting: The advisor stays current with global AI regulations, guidance and industry standards. Exercise: Practice summarizing a current AI governance development in terms of its scope, product implications and questions requiring further review; the posting names no particular framework.

Skills to revise

• Document an independent ethics review

  Posting: The advisor independently conducts and documents reviews of new and existing AI projects. Exercise: Draft a short review for a hypothetical consumer-facing AI feature covering intended use, potential harms, evidence, unresolved questions and recommended follow-up.

• Design risk-based tests

  Posting: Testing and red-teaming are used to assess unfair treatment, safety and misuse risks. Exercise: Write a small set of test scenarios for the hypothetical feature, stating the harm each probes, what to record and how findings would inform the review.

• Advise across product teams

  Posting: The role provides sustained advice to product and business teams and collaborates with technical teams. Exercise: Rehearse a brief discussion that translates a review finding into a product decision, identifies the evidence needed and addresses differing stakeholder priorities.

• Improve a review process

  Posting: The advisor improves review processes, documentation, tools and criteria. Exercise: Critique a sample review template you create, then revise its criteria to make harm identification, evidence and follow-up more consistent.

• Teach Responsible AI by design

  Posting: The advisor develops and delivers internal training and works with the Inclusive Product Design Office. Exercise: Draft a five-minute training outline using one consumer-product risk example and a prompt for teams to consider affected users early in design.

• Map relevant experience

  Prepare a one-page evidence map covering the stated 8+ years of relevant experience and examples of independent judgment, AI assurance, and cross-functional work. The degree-or-equivalent-experience requirement is source-backed; the map is a preparation aid.

• Build an AI Ethics review walkthrough

  Select a review you can discuss and prepare a concise account of its scope, evaluation or testing methods, documented findings, recommendations, and follow-up. The role calls for reviews of both new and existing AI projects.

• Practice a consumer AI risk assessment

  For a hypothetical consumer-facing generative or agentic feature, draft a review outline covering unfair treatment, transparency, safety, misuse, and human oversight. Include proposed tests and monitoring; this is a practice exercise, not a stated Lenovo project.

• Prepare a governance reference sheet

  Summarize a few relevant AI regulations, guidance documents, or standards and how you would check their applicability during a review. The posting calls for staying current globally but does not name a required framework.

• Show how you would improve a review process

  Bring a short example of a review criterion, documentation template, or workflow improvement, including how you would seek input from product and technical teams. Lenovo specifically names process improvement and partnership with its Inclusive Product Design Office.

• Outline an internal training segment

  Draft a brief Responsible AI by design session for product teams, with a learning objective, a consumer-product example, and a practical takeaway. Developing and delivering internal training is part of the role.

• Write questions for the interviewers

  Prepare questions about the assigned consumer-product portfolio, how reviews continue through deployment and monitoring, how teams work with the Inclusive Product Design Office, and where review criteria most need improvement.