Anthropic announced Claude Science, a new AI workbench designed to help scientists consolidate fragmented tools and datasets into a single environment. The platform, unveiled at an event called "The Briefing: AI for Science" this week, generates figures and visuals to support research workflows. Anthropic framed the launch around drug development, signalling an expansion beyond its core business in coding tools and large language models into the life sciences sector. Source
Midjourney, the AI startup known for image generation, released a behind-the-scenes video of its dunk-tank ultrasound scanner, a device it plans to deploy in spas and position as a transformation for medical imaging through cheap, detailed, radiation-free scanning. The nearly 20-minute tour offered more detail on the futuristic device, though the company has yet to demonstrate substantial proof of its efficacy. Source
Unionization negotiations at Google DeepMind got off to a difficult start on Wednesday, with employees expressing frustration about what they saw as executive unwillingness to engage meaningfully with the prospect of unionization. The tense opening exchanges suggest the talks face significant obstacles as workers seek formal representation at the AI research unit. Source
Microsoft disclosed tax planning tactics previously used in secrecy, providing a rare window into how major technology companies manage their tax obligations through international structures. The revelation comes as a new European directive will soon require other U.S. companies to provide similar disclosures, marking a shift toward greater transparency in corporate tax reporting. Source
Britain's National Crime Agency warned parents against publicly sharing images of their children online, citing a growing threat of synthetic child abuse material being created with AI tools. The alert reflects rising concern among law enforcement about how publicly available photographs can be misused by bad actors to generate illegal content. Source