The Saudi Times

Cybersecurity firm Outpost24 has released an AI-powered penetration testing tool designed to help organizations identify vulnerabilities in their systems at a pace that matches the speed at which AI is being deployed

General Compute has announced the general availability of its ASIC-based inference cloud platform, purpose-built for AI agent workloads. The service, which opens to the public on May 15, uses custom silicon accelerators

xAI has released Grok 4.3 in beta, an incremental but meaningful update to its flagship conversational AI model. The release brings a revised model architecture, improved reasoning consistency, and a knowledge base

Mozilla has introduced Thunderbolt, an open-source AI client designed to give individuals and organizations full control over their AI interactions by enabling self-hosting. The release marks Mozilla’s most significant move yet into

OpenAI has introduced GPT-Rosalind, a purpose-built reasoning model designed to accelerate pharmaceutical research, genomics analysis, and molecular biology. The model, named after pioneering scientist Rosalind Franklin, is already being integrated by Amgen,

Google has introduced TurboQuant, a new quantization framework that reduces the memory footprint of large language models by a factor of six while preserving performance at or near frontier levels. The breakthrough

Anthropic has unveiled Claude Mythos 5, its most powerful large language model to date, featuring an estimated 10 trillion parameters. The release marks a significant leap in scale and capability, with the

In a notable cybersecurity development, recent reports indicate that vulnerability discovery is no longer limited to the latest artificial intelligence models. An older model, known as “Opus,” has reportedly succeeded in developing