OpenAI launches Presence for enterprise agents ↗
OpenAI introduced Presence, a managed product for deploying voice and chat agents across customer support, sales and internal workflows. Companies control permissions, policies, approvals and the point at which a human needs to step in - sensible, frankly.
OpenAI says its own support agent resolves 75% of inbound issues without human help. Presence is initially limited to eligible enterprise customers, with deployments managed by specialist engineers and selected partners.
OpenAI’s planned cloud spending hits $750 billion ↗
OpenAI’s projected computing commitments have reportedly risen from roughly $600 billion to around $750 billion. AI infrastructure spending is becoming less of a budget line and more of a small moon.
The plan includes larger cloud agreements and OpenAI’s first directly developed data-centre project, a vast Georgia facility requiring 3.2 gigawatts of power. Ambitious... or slightly bonkers, perhaps both.
AMD to sell Anthropic tens of billions in AI servers and invest up to $5 billion ↗
AMD agreed to supply Anthropic with tens of billions of dollars’ worth of AI servers while investing as much as $5 billion in the Claude maker. The investment depends on Anthropic reaching deployment milestones.
Anthropic plans to acquire up to two gigawatts of AMD’s Instinct MI450 chips. It gives Anthropic some desperately needed compute and hands AMD a surprisingly substantial win against Nvidia.
US to spend $5 billion on research powered by AI ↗
The US government plans to commit $5 billion to AI-powered scientific research involving 15 federal agencies. Researchers will gain access to government supercomputers, specialist datasets and other experimental infrastructure.
Projects will examine chronic diseases, accelerate drug discovery and develop longer-lasting construction materials. It is AI policy becoming unusually practical - medicine, concrete, the whole toolbox.
Google increases spending forecast after cloud growth surges ↗
Google Cloud revenue jumped 82% to $24.8 billion as enterprise demand for AI computing accelerated. Alphabet also raised its capital-spending forecast to between $195 billion and $205 billion.
Investors were less thrilled, perhaps unsurprisingly. Alphabet reported negative free cash flow, shares slipped and executives faced questions about delays to Gemini 3.5 Pro, though training for Gemini 4 has begun.
Amazon cuts jobs inside its AGI group ↗
Amazon eliminated roles within parts of its artificial general intelligence organisation. The company said it was narrowing its focus to the initiatives that matter most to customers, which is corporate language with fairly sharp teeth.
The full scale of the cuts remains unclear. They follow leadership departures and a restructuring that placed AGI work alongside Amazon’s chip and quantum-computing operations.
FAQ
What are the biggest AI news developments this week?
The main developments include OpenAI’s launch of Presence for enterprise agents, growing infrastructure commitments, and major investment in AI computing. AMD is supplying Anthropic with servers, while Google has raised its spending forecast. The US government is also funding AI-powered research, as Amazon reduces roles across parts of its AGI organisation.
What is OpenAI Presence and how will companies use it?
OpenAI Presence is a managed product designed to deploy voice and chat agents across customer support, sales and internal workflows. Businesses can define permissions, policies, approval requirements and rules for human intervention. It is initially available to eligible enterprise customers, with deployments supported by specialist engineers and selected partners.
Why is OpenAI planning to spend so much on AI infrastructure?
OpenAI reportedly expects its computing commitments to reach around $750 billion as demand for model training and deployment continues to grow. Its plans include larger cloud agreements and a directly developed data-centre project in Georgia. The facility is expected to require substantial power capacity, underlining how compute and energy have become central constraints in advanced AI development.
How does the AMD and Anthropic agreement affect the AI chip market?
AMD will supply Anthropic with tens of billions of dollars’ worth of AI servers and may invest up to $5 billion in the company. Anthropic plans to acquire up to two gigawatts of AMD Instinct MI450 chips. The agreement gives Anthropic additional computing capacity while providing AMD with a significant opportunity to compete more directly with Nvidia.
What will the US government’s $5 billion AI research programme support?
The programme will support AI-powered scientific research across 15 federal agencies. Researchers are expected to gain access to government supercomputers, specialist datasets and experimental infrastructure. Planned projects include studying chronic diseases, accelerating drug discovery and developing longer-lasting construction materials.
Why are AI companies increasing spending while also cutting jobs?
AI news increasingly shows companies concentrating resources on computing infrastructure and selected high-priority projects. Google has raised its capital-spending forecast as cloud demand grows, while Amazon is narrowing parts of its AGI organisation. This suggests companies may continue investing heavily in chips, data centres and customer-facing products while restructuring teams that do not align with immediate priorities.