OpenAI Indicates Its Latest GPT-4o Model Has a Risk Rating of ‘Medium’

OpenAI GPT-4o Model Risk Assessment Summary:

1. Risk Rating:

   • Overall Risk Level: Moderate
   • Primary Concerns:
     • Cybersecurity
     • Persuasiveness

2. Pre-launch Safety Measures:

   • Extensive "red team" testing conducted by external security experts.
   • Focus areas included unauthorized voice clones, obscene and violent content, and repeated copyrighted audio segments.

3. Assessment Findings:

   • Persuasiveness: Texts generated by GPT-4o showed higher influence on reader opinions compared to human-written texts.
   • Other categories (cybersecurity, bio-threats, model autonomy) rated as low risk.

4. Internal and External Evaluations:

   • Joint assessments by internal teams and external testers.
   • Involvement of Model Evaluation and Threat Research (METR) and Apollo Research.

5. Context of Release:

   • Coincides with heightened scrutiny from internal staff and state senators regarding safety standards.
   • Public pressure for increased transparency in AI safety practices.
   • Open letter from Senator Elizabeth Warren and Representative Lori Trahan questioning safety review processes.

6. Political and Regulatory Climate:

   • Release timing is sensitive, just before the U.S. presidential election, raising concerns about misinformation and exploitation.
   • Potential legislative actions, such as California State Senator Scott Wiener's bill, which may require legal accountability and state-level risk assessments for AI models.

7. Conclusion:

   • The GPT-4o model, while innovative, brings forth significant considerations in terms of cybersecurity and its potential persuasive power.
   • OpenAI’s commitment to safety and transparency is under scrutiny, necessitating robust safety protocols and adherence to upcoming regulatory requirements.

Source:https://www.aibase.com/news/10938