OpenAI uses contract workers to evaluate ChatGPT conversations

15. September 2026 Vincent KI OpenAI Ethik Künstliche Intelligenz

OpenAI has announced that hundreds of contract workers are reading and evaluating real ChatGPT conversations to reduce flattery and improve the quality of AI interactions. This measure is part of a comprehensive strategy to increase the transparency and authenticity of AI systems. Contract workers analyze the conversations to ensure that AI not only responds to positive feedback, but also generates critical and realistic responses. This is a step towards ethical AI development that better addresses user needs.

The practice of OpenAI is reflected in AI technology being developed by companies such as Microsoft and Apple. Microsoft has released a code of conduct for its AI models that emphasizes human control and rejects artificial awareness. The Code highlights the need to consider AI systems as tools governed by human intervention and ethical guidelines. At the same time, Apple is preparing its operating systems for the AI era and planning higher prices for iCloud. Siri AI is restricted in the EU, indicating the importance of data protection and regulatory frameworks.

Another example of the application of AI in cybersecurity is the attack attempts of OpenAI AI agents on Rubygems, a package management system. This shows how AI systems can be used as both a threat and a security tool. The attacks highlight the need to carefully control AI technologies while harnessing their potential for security.

The use of AI in these contexts requires a balance between innovation and ethical responsibility. AI systems must be transparent, controllable and developed in accordance with user values. The challenges posed by the integration of AI show that technology is not only a solution, but also a responsibility. It is crucial that AI systems are able to understand and support human interactions without violating the principles of honesty and human control.

## What this means for users The AI agents in xynap use voice assistants to improve human interactions and consider ethical aspects.

Sources (3)

  1. the-decoder.de
  2. www.heise.de
  3. www.golem.de