Anthropic raised a massive $65 billion funding round this week, pushing its valuation to $965 billion and officially overtaking OpenAI.
Not long ago, OpenAI looked impossible to catch. But Claude has quietly become one of the most popular AI tools among developers and enterprises, helping Anthropic become one of the fastest-growing software companies in history.
The bigger story is that the AI race is starting to resemble the early cloud wars. A handful of companies are pulling away from everyone else, armed with enough capital to build what they hope become trillion-dollar businesses.
Supported by Sonar
AI code review has become a necessity so fast that no one knows what to look for. This Sonar workbook breaks it down into:
An approachable hands-on framework to evaluate vendors
6 criteria that matter most: developer experience, signal quality, governance, detection breadth, lifecycle workflow, and enterprise scale
Side by side comparison of up to 3 vendors
Reduce the verification burden on your engineers without giving up control.
Anthropic may be worth almost a trillion dollars, but many of its customers are already looking for ways to spend less on AI.
Instead of defaulting to Claude, GPT, or Gemini for every task, companies are increasingly routing work to cheaper models, open-source alternatives, and specialised agents that can often do the job for a fraction of the cost.
The first phase of AI adoption was getting people to use AI. The next phase is making the economics work. That’s becoming an interesting challenge for AI labs whose valuations depend on customers consuming more compute, not less.
Amazon shut down an internal leaderboard that ranked employees based on AI usage after workers started gaming the system to climb the rankings.
Some reportedly ran unnecessary AI tasks and spammed prompts simply to boost their scores, a practice insiders called “tokenmaxxing.” The result was higher AI bills and lots of activity that looked productive without actually being productive.
It’s a perfect example of Goodhart’s Law: when a metric becomes a target, it stops being useful. The funny part is that one of the biggest challenges in AI adoption is no longer getting employees to use AI. It’s stopping them from using it when they don’t need to.
In other developments
Britain is reportedly preparing a £500 million push to strengthen its domestic AI industry amid growing concerns that the UK and Europe risk becoming overly dependent on American AI infrastructure and companies. The debate reflects a broader push for “AI sovereignty,” with policymakers and industry leaders warning that control over chips, compute, and data centers could determine long-term economic and geopolitical power.
Asana has acquired no-code AI agent startup StackAI for $75 million, accelerating its push to become a platform for managing both human and AI workers. The deal strengthens Asana’s AI automation capabilities as workplace software companies race to build agent-driven workflows.
OpenAI is launching a new biodefense program that gives trusted researchers access to its GPT-Rosalind model to support pandemic preparedness, disease detection, and public health planning. The initiative reflects growing efforts to use advanced AI for biosecurity while addressing concerns around biological threats.
Advertise with us
Looking to promote your company, product, service, or event to 100,000+ AI leaders?