|
Reuters
OpenAI has reportedly signed a seven-year, $38bn deal with Amazon Web Services to access "hundreds of thousands" of Nvidia GPUs. The move loosens Microsoft's exclusive grip on OpenAI's workloads and signals intensifying cloud competition for frontier model training. Strategic Insight
Multi-cloud AI infrastructure is becoming standard for leading AI companies, reducing vendor lock-in and ensuring capacity during peak demand periods. |
|
TechCrunch
Anthropic has raised its internal targets, according to reporting, with steep ARR ambitions over the next three years as enterprise adoption accelerates. While not formal guidance, the figures reinforce expectations that spending on training compute and safety research will continue to climb. Strategic Insight
Enterprise AI adoption is accelerating beyond early-adopter phase, with companies prioritising safety-focused AI providers that can scale with their needs. |
|
anthropic.com
Iceland's Ministry of Education and Children is partnering with Anthropic to provide teachers nationwide with access to Claude as part of a comprehensive AI education pilot. The programme will test how AI can support lesson planning and personalised learning across urban and remote schools. Strategic Insight
National-level AI education initiatives signal growing government recognition of AI as essential infrastructure, with implications for teacher training and curriculum development. |
|
GeekWire
The Allen Institute for AI introduced "OlmoEarth," a set of open-source models that analyse satellite data to map wildfire risk and other environmental signals. The effort aims to reduce dependence on Big Tech platforms for geospatial AI while enabling researchers and NGOs to run models locally. Strategic Insight
Open-source AI for environmental monitoring democratises access to climate intelligence, particularly valuable for organisations with limited budgets or data sovereignty concerns. |
|
The Verge
A Verge analysis argues that new "agent" tools, which can complete tasks across the web, are making it easier for students to cheat — and that some AI companies deflect responsibility. The piece surveys current agents' capabilities and the policy grey areas schools are grappling with. Strategic Insight
Educational institutions must urgently rethink assessment methods as agentic AI blurs the line between assistance and automation, requiring new frameworks for evaluating student learning. |