The tech news that matters, without the noise.

AI is entering a new phase.

It isn't just answering questions or generating content anymore. The big tech companies are building AI that can work on your behalf — while simultaneously spending hundreds of billions of dollars building the infrastructure to make that possible.

Here are three stories worth knowing today.

OpenAI has introduced dots, its new always-on AI agents.

Unlike a chatbot that waits for your next question, a dot can take a project and keep working on it using its own cloud computer and browser. OpenAI says dots can connect to more than 4,000 apps and can work toward goals over time. Users can set permissions and decide when approval is required.

OpenAI gives examples including investigating bugs, updating launch materials, analysing new information and preparing content for approval.

This is a big shift.

The old AI model was:

You ask → AI answers → you decide what to do next.

The new model is closer to:

You give AI the goal → AI works → you review the result.

That could save enormous amounts of time — but it also means permissions, privacy and checking the work become much more important.

AI isn't just becoming an assistant. It's becoming a worker.

Anthropic's IPO filing has revealed just how expensive the AI race is becoming.

According to Reuters, Anthropic plans at least $518 billion in cloud, computing and infrastructure commitments over the coming decade. About 80% of those commitments are reportedly non-cancelable or require payment regardless of usage.

The commitments include more than $111 billion with Google, $110 billion with Amazon and $31.4 billion with Microsoft, alongside major equipment obligations with Broadcom.

Anthropic's revenue grew sharply in 2025, but the company also reported a $42 billion net loss, including a large accounting charge.

We keep hearing about amazing AI models.

But behind those models is something much less glamorous:

data centres, chips, electricity, cloud contracts and billions upon billions of dollars.

The AI race isn't just a software race anymore.

It's an infrastructure race.

And someone eventually has to pay the bill.

Meta has launched Meta Enterprise Platform, a new business focused on bringing its AI models, agents and developer tools directly into companies.

The platform includes tools such as Meta's Muse agent, Meta Business Agent, Muse API and Muse Code.

Meta has also hired former MongoDB CEO CJ Desai to lead the new enterprise push.

Meta has spent years building platforms where people connect, communicate and advertise.

Now it's trying to become part of how businesses actually work.

That puts Meta into the same much bigger race we're seeing across the industry:

Who becomes the AI layer sitting inside the workplace?

Because if AI becomes something companies rely on every day, controlling that layer could be a very big business.

Three different stories.

One bigger trend:

AI is moving from something you use to something that works alongside you — and increasingly, for you.

The technology is advancing quickly.

But so are the questions around cost, control, privacy and responsibility.

☑️ OpenAI has introduced always-on AI agents called dots.
☑️ Anthropic has disclosed $518 billion in planned infrastructure commitments.
☑️ Meta is pushing deeper into enterprise AI.
☑️ The next phase of AI may be less about answering questions — and more about getting things done.

That's your OneClick.

See you tomorrow. 💎