Anthropic
AI company developing Claude AI assistant
Every fact below is drawn from Via News's knowledge graph and linked to the source document it came from. Nothing here is unsourced. How we source →
Key metrics · each point sourced
Stated objectives
Complete $50 billion US data center buildout
sourceProve that superior orchestration of existing models can yield exponential productivity gains as alternative to trillion-dollar infrastructure build-outs
sourceHalt training or delete model weights if safety and security standards cannot be met
sourceTarget users running Claude Code continuously in the background 24/7 with new rate limits
sourceRace to develop expert-level and superhuman AI systems
sourceExpand Cowork beyond macOS to Windows and add cross-device sync, growing access beyond Claude Max subscribers
sourceMake working with Claude feel less like operating a tool and more like delegating to a colleague, targeting mainstream non-technical users
sourceContinue developing agent safety mechanisms including defenses against prompt injection attacks as an active area of development
sourceAnthropic reportedly targeting its own IPO
sourceDiversify compute supply beyond Nvidia by expanding use of Amazon Trainium and Google TPUs to reduce concentration risk and improve hardware availability
source