OshenDavidson

Scroll to explore
Get to know me

AI Safety Researcher

Hi! I'm Oshen

I’m interested in how AI systems communicate, how people interpret their responses, and how those interactions shape human autonomy, beliefs, trust, and decisions. That includes questions about misleading explanations, how uncertainty is presented, and whether AI systems genuinely support humans in exercising their own judgment.

My research experience spans observable model behavior, including areas such as: information integrity, AI search engines’ influence on consumer behavior, persistence in output generation, model evaluations, and more. My background is in designing, leading and executing empirical end-to-end research.

Throughout my work, my main goal is to contribute to the field’s understanding of AI as its capabilities grow, helping society anticipate risks, develop stronger safeguards and defenses, and make informed decisions about these systems. Additional interests in the field include model psychology and AI welfare, particularly how we interpret behavior when considering questions about agency, experience, and moral status.

Outside of research, you can find me watching Jeopardy, getting absorbed in a new hobby, tinkering with side projects, or reading.

More about me →

Why AI Safety Matters to Me

As AI capabilities increase and systems operate with more autonomy, our efforts to understand model behavior need to scale alongside them. This means expanding interpretability research, examining capabilities and limitations, and identifying vulnerabilities in existing defenses. This also means strengthening our ability to anticipate emerging risks and develop safeguards to address them.

That understanding also needs to account for AI's influence on people and society. How a model presents information, expresses uncertainty, and responds to us can affect belief formation, trust, and human autonomy. Building effective safeguards depends on both understanding the depth of these systems' capabilities, as well as how these effects can impact individual behavior.

Through empirical studies and behavioral evaluations, my work is one contribution to a shared effort to better understand AI. That effort grows stronger as people share knowledge, build on one another's work, and bring different perspectives to the same challenges. If you're thinking about these questions too, I'd love to connect, compare notes, or collaborate on a study.

Specimen 001 · citation behavior
Citation Selection in Commercial AI Search: A Stage-Based Analysis of Retrieved Pages
Collected Apr 2026
Read the study →
Specimen 002 · citation behavior
From Search to Citation: Measuring How ChatGPT Retrieves and Selects Sources
Collected Mar 2026
Read the study →
Specimen 003 · ugc
Community Sources in AI Search: A Cross-Model Analysis of 5.5 Million Answers
Collected Nov 2025
Read the study →

Research Coverage & Citations

  • Zyppy
  • Search Engine Land
  • MSN
  • eMarketer
  • MSN Deutschland
  • Forrester