Internship at Volkswagen Group Innovation (f/m/d)

Role overview

Join Volkswagen Group Innovation in Wolfsburg and contribute to a team working with modern approaches to automotive design and future mobility technologies. The work is carried out in an interdisciplinary setting, with innovation hubs in Europe, Asia, and the United States that combine local insight with global thinking.

During the internship, you will collaborate on a variety of mobility-related topics and strengthen both your technical and personal capabilities. The core assignment is to study how AI agents perform and how robust they remain when exposed to different attack methods.

What you will work on

• Build a workflow for creating adversarial examples for long-horizon datasets.
• Use both quantitative and qualitative approaches to assess the output of the pipeline.
• Design, test, and compare agents facing different adversarial attack styles across multiple agentic AI architectures.
• Examine ways to measure the performance and resilience of different agents.

Profile and qualifications

The company is looking for a driven and hands-on candidate with a keen interest in computational linguistics and its use in agent-based AI systems.

• You should be enrolled in Computer Science, Data Science, Computational Linguistics, or a similar degree program.
• Strong Python coding ability is required.
• You need a solid understanding of generative AI, agentic AI concepts, and their practical applications.
• You should bring strong analytical thinking, good problem-solving ability, and close attention to detail.
• German and English should both be at B2 level.

Application documents

• CV or résumé.
• Current certificate of enrollment, unless you are applying for a gap year.
• Latest transcript of records.
• For a mandatory internship, a university confirmation letter is also needed.
• Non-EU applicants must provide a valid work permit.

Keywords

Agentic AI, robust agentic systems, adversarial attacks, AI evaluation metrics.