Records found: 6
'Judges are piloting generative AI to speed research and drafting, but hallucinations and weak accountability have already produced flawed orders and raised urgent questions about safe use.'
OpenAI’s new GPT-5 model offers faster reasoning, better user experience, and fewer hallucinations, but represents a refinement rather than a breakthrough on the path to AGI.
Hirundo raises $8 million to develop machine unlearning technology that removes AI hallucinations and biases, offering enterprises a more reliable and efficient way to improve AI model safety.
Researchers developed the SUM dataset to teach AI language models to say 'I don't know,' significantly reducing hallucinations and improving refusal rates without harming accuracy on answerable problems.
AI feedback loops occur when AI models train on outputs from other AI systems, causing errors to compound and potentially leading to serious business risks. Understanding and mitigating these loops is critical for safe AI deployment.
Patronus AI’s Judge-Image leverages Google Gemini to enhance the evaluation of multimodal AI systems, ensuring accurate image-to-text outputs and setting new standards for AI reliability.