Asking an AI for salary negotiation advice is a matter of concern: Controlled experimental perturbation of ChatGPT for protected and non-protected group discrimination on a contextual task with no clear ground truth answers

Synopsis

Audit of four ChatGPT versions across ~400k prompts varying gender / university / major; finds significant inter-version inconsistency (~40% rise in recommendations Jun 2023→May 2024), gender gaps, and elite-university and STEM premia, plus reasonable salaries even for fictional universities.

Keywords

·ChatGPT bias ·salary negotiation ·version comparison ·gender bias ·prompt audit

Related entries

• Dynamic calibration of trust and trustworthiness in AI-enabled systems
  2026 · International Journal on Software Tools for Technology Transfer 2026
• Trust in human-AI collaboration in finance: a bibliometric-systematic literature review
  May 8, 2026 · AI & SOCIETY 2026
• A Co-Evolutionary Theory of Human-AI Coexistence: Mutualism, Governance, and Dynamics in Complex Societies
  April 24, 2026 · arXiv
• Personality and Personal AI Agents: A Co-Evolutionary Framework (PACE)
  March 7, 2026 · International Journal on Social and Education Sciences 2026
• Boosting Metacognition in Entangled Human-AI Interaction to Navigate Cognitive-Behavioral Drift
  February 2, 2026 · arXiv
• Human-AI perception: not much different, but some distinct novelties
  January 29, 2026 · The Bottom Line 2026