Science

New research demonstrates how combining artificial intelligence with vulnerability data can significantly improve the prediction of zero-day exploit severity.

A new approach allows engineers to translate imprecise, natural language descriptions of system behavior into a formal, verifiable logic.

Researchers have developed a novel watermarking technique that embeds a hidden signature within deep neural networks, making it harder for malicious actors to steal or repurpose AI models.

A novel algorithm addresses the challenges of training graph neural networks on fragmented, diverse datasets without compromising performance or generalization.

Researchers have developed a new method leveraging the power of large language models to automatically identify and analyze the human values expressed within written content.

A novel architectural approach embeds societal values directly into AI design, enabling continuous oversight and adaptation of complex socio-technical behaviors.

Researchers have developed a highly effective phishing detection system leveraging the power of character-level analysis and advanced machine learning techniques.

A new hybrid modeling approach combines the accuracy of physics-based simulations with the adaptability of data-driven machine learning to overcome limitations in complex system modeling.

As disinformation becomes increasingly sophisticated and commercially available, a coordinated, multidisciplinary approach is crucial to defend against large-scale attacks on public trust.

As artificial intelligence increasingly steps into the role of mental health support, ensuring users understand how and why AI arrives at its conclusions is paramount.