OpenAI lässt externe Gruppen KI-Modelle früher prüfen

33 Quellen
  • OpenAI gab am Dienstag bekannt, dass externe Organisationen Sicherheitsbewertungen von KI-Modellen bereits in früheren Entwicklungsphasen durchführen dürfen, so Bloomberg.
  • Der Schritt folgt auf die Forderung des Konkurrenten Anthropic nach eingebetteten Drittprüfern und den Rücktritt eines Forschers, der warnte, dass KI-Labore "mit unserem Leben spielen".
  • Kritiker bezweifeln, dass eine sinnvolle Aufsicht möglich ist, während die Labore auf Billionen-Bewertungen und über 7 Billionen Dollar an geplanten KI-Investitionen zusteuern, berichtet Axios.
Quellen (33)
  1. 1 OpenAI to Let Outside Groups Evaluate AI Models at Earlier Phase www.bloomberg.com
  2. 2 OpenAI proposes development of global AI standards for ... www.cnbc.com
  3. 3 Anthropic Researcher Resigns: Jacob Coxon on the Race (Sept ... www.explainx.ai
  4. 4 With $7 trillion on the line, AI safety may always take a backseat to market domination www.axios.com
  5. 5 Anthropic and OpenAI need independent safety evaluators, experts ... www.cnbc.com
  6. 6 Our framework for reporting model misalignment - OpenAI openai.com
  7. 7 Pacing model development in an era of cyber-critical capabilities openai.com
  8. 8 Path to Astra: critical capabilities and frontier safeguards - OpenAI openai.com
  9. 9 The AI policy window is open. We need to act. | OpenAI openai.com
  10. 10 September 2026's AI Safety Incident Cluster - CASRAI casrai.org
  11. 11 Research acceleration: The view inside OpenAI openai.com
  12. 12 From hard refusals to safe-completions: toward output-centric safety ... openai.com
  13. 13 OpenAI pushes for mandatory national AI safety rules - Reuters www.reuters.com
  14. 14 OpenAI Pauses Training: July to August 2026 - AllyNerds www.allynerds.com
  15. 15 Third-party cyber evaluations involving OpenAI models openai.com
  16. 16 Is OpenAI Intentionally Distancing Itself from Safety? - Annielytics.com www.annielytics.com
  17. 17 OpenAI Calls for Mandatory AI Safety Rules | EM360Tech em360tech.com
  18. 18 Safety and alignment in an era of long-horizon models - OpenAI openai.com
  19. 19 OpenAI reveals six more safety issues and unveils plan to disclose ... www.bbc.com
  20. 20 OpenAI | Research & Deployment openai.com
  21. 21 ChatGPT: Chat, Work, Create & Code with AI chatgpt.com
  22. 22 Anthropic Researcher Jacob Coxon Resigns, Warns AI Industry ... deadline.com
  23. 23 GPT-4 | OpenAI openai.com
  24. 24 OpenAI and Anthropic Researchers Are Warning About AI Risks time.com
  25. 25 OpenAI - Wikipedia en.wikipedia.org
  26. 26 Jacob Coxon’s Resignation From Anthropic Explodes the AI ... political.org
  27. 27 OpenAI Platform platform.openai.com
  28. 28 He Helped Build Powerful AI at OpenAI and Anthropic ... - TIME time.com
  29. 29 About - OpenAI openai.com
  30. 30 An Anthropic researcher just quit, saying OpenAI and ... tech.yahoo.com
  31. 31 API Platform | OpenAI openai.com
  32. 32 Anthropic researcher quits, warns AI could wipe out humanity ... www.yahoo.com
  33. 33 GPT‑5 is here - OpenAI openai.com