Legacy Code

After weeks of negotiations with the Trump administration, Anthropic has received approval to restore access to its most advanced Mythos AI model, though only to a carefully curated group of US companies and government agencies. The decision ends a period in which the model was completely unavailable, during which the company worked through regulatory concerns with federal officials.

ElevenLabs is scaling its hyper-realistic voice synthesis technology across Latin American markets, expanding availability of its AI voice generation platform in a region with significant untapped demand for voice-based applications. The expansion includes localization efforts for Spanish and Portuguese language variants with regional accent and dialect support.

European leaders are accelerating efforts to develop indigenous AI capabilities rather than rely on American and Chinese models, citing both strategic independence concerns and dissatisfaction with current options. The continent's leverage has increased as US policy uncertainty creates openings for alternative centers of AI development.

Meta has unveiled an AI workforce academy designed to equip employees and partners with AI skills as the company accelerates its AI-driven product roadmap. The initiative has garnered public support from former Trump administration officials and business leaders, who view it as a model other companies should emulate.

OpenAI has postponed the rollout of its GPT-5.6 model following a request from the White House, marking the second major AI capability release to face government intervention in as many weeks. The delay comes shortly after Anthropic was forced to temporarily take its advanced Mythos model offline under similar pressure.

OpenAI has announced GPT-5.6 Sol, a next-generation model demonstrating significant improvements in coding, scientific reasoning, and cybersecurity applications. The model incorporates OpenAI's most comprehensive safety framework to date, designed to address ongoing concerns about frontier AI deployment.

Paul Meade, the Apple vice president who led the Vision Pro hardware initiative, has reportedly departed Apple to join OpenAI's hardware engineering team. The move signals OpenAI's commitment to building proprietary hardware devices to deliver its AI models to consumers and enterprises.

Elon Musk's proposal to build data centers in orbit has drawn skepticism from prominent technology investors and executives, including SoftBank's leadership, who question the technical feasibility and economic viability of the concept. Critics argue that orbital infrastructure presents enormous engineering challenges and operational costs relative to terrestrial alternatives.