THE WORLD IS NOT STANDING STILL.RSS
BIG CHANGE.
BIG CHANGE, one tap away
BIG CHANGE, one tap away

Make room for the big picture.

Add BIG CHANGE to your home screen or desktop. It opens in its own window, with the latest stories when you’re online.

iPhone & iPad
Open bigchange.ai in Safari. Tap Share, then Add to Home Screen. Turn on Open as Web App if shown, then tap Add.
Android
Open bigchange.ai in Chrome. In the browser menu, choose Install app or Add to Home screen.
Desktop
In Chrome or Edge, use the install icon in the address bar or the browser’s app menu. In Safari on Mac, choose File → Add to Dock.
All briefings

Follow the change ·

The day,
in perspective.

What happened in AI and robotics — and the change behind it. A reading path through 32 stories from BIG CHANGE.

Ordered using RADAR attention, story opens and likes. Excerpts from our published reporting; open a story for the full context and sources. Dates in UTC.

news

SecurityScorecard’s benchmark tests AI on vendor-risk decisions

A vendor-risk benchmark evaluates AI task performance and resistance to manipulation.

The big change

The tasks turn risk decisions into questions with verifiable answers.

It separates task performance from resistance to manipulation. That matters when an AI system reads supplier-supplied material: the content can be relevant evidence while also containing instructions or claims that should not determine the answer.

A score belongs to its task set and run conditions. Before using it to select or oversee a tool, a team still needs evidence that the result holds for its own data, workflow and controls. A leaderboard alone cannot supply that assurance.

Read the full story
—
Back to the briefing
news

Chrome begins using Gmail context for enhanced autofill

Chrome is rolling out Gmail-based autofill suggestions to eligible US Gemini users, who can control access in settings.

The big change

Gmail details can appear beside an online form, saving a trip to the inbox.

The suggested value still needs to match the transaction and the receiving site; a plausible-looking reference can belong to another booking or order.

In the US, Gemini in Chrome users aged 18 or older need Personal Intelligence; Photos access is planned later.

Read the full story
—
Back to the briefing
news

Google plans restricted Gemini 4 Carbon access for cyber defenders

Google plans initial Carbon access for selected cyber defenders, with benchmark results and introductory prices announced.

The big change

Google plans a phased Fairwind rollout, starting with a subset of cyber defenders. Most teams must wait before testing the model themselves.

Artificial Analysis's lead is tied to its measured configuration and mix of tasks; a buyer still needs evidence from its own workload.

Google reports an 88.2% result on the vulnerability-finding SEC-bench Pro without guardrails. Teams need to assess the safeguards and behavior of the version they can obtain.

Read the full story
—
Back to the briefing
news

GitHub Enterprise adds metered Copilot access

GitHub Enterprise’s new Copilot access has metered AI use, with spending limits that require enforcement settings.

The big change

Copilot access can reach eligible Enterprise Cloud users without a separate Copilot seat charge, but that does not turn AI work into a fixed-price benefit. The purchasing question moves toward how much work people run and which features they use.

A spending control can serve very different purposes: warning an administrator or stopping requests. That choice connects financial oversight to everyday availability, so the person setting the budget needs to understand the effect on the people doing the work.

Procurement teams comparing the options need a view of total expected consumption as well as license costs. An estimate built from the organization’s own usage and purchasing records can make that choice more informative than a comparison based on headcount alone.

Read the full story
—
Back to the briefing
news

How CUDA 13.4 enforces shared GPU memory in Kubernetes

Configured KAI Scheduler and GPU Fractioning components connect CUDA 13.4 memory limits to shared Kubernetes GPU workloads.

The big change

A fractional GPU request becomes a runtime boundary only when scheduling, the operator and a supported node stack are configured together. Placement accounts for the requested share; the node applies the limit to the container.

Memory limits and compute-sharing mode answer different operational questions. A team choosing a mode must consider how a workload failure affects its neighbors.

A workload may be allowed to borrow unused memory, but its recovery when that memory is reclaimed matters as much as utilization. The documented CUDA 13.4 and driver r615 limitations qualify the promised minimum in specific Run:ai reclaim cases.

Read the full story
—
Back to the briefing
news

Google Cloud revenue doubles in third quarter

Cloud growth and company cash flow.

The big change

Cloud revenue does not isolate any one AI product’s contribution. Investors therefore cannot infer the return on a specific model from the segment total.

Segment operating income and company-wide free cash flow answer different questions. A profitable segment can coexist with negative company-wide free cash flow. Asset purchases use cash when paid for, while depreciation generally spreads their cost over time.

The timing and funding of investment matter to shareholders as well as the growth rate. Borrowing and share issuance provide financing, while additional shares can dilute existing holdings; the eventual return depends on cash generated over the assets’ useful lives.

Read the full story
—
Back to the briefing
news

Cloudera launches Blueprints, plans NVIDIA workflows

Cloudera launched Blueprints and described NVIDIA workflows as forthcoming.

The big change

Reference applications give data teams a starting point for building AI systems, but the work of connecting them to an organization’s own data and controls remains. A reusable pattern can make an initial evaluation more concrete without settling whether it fits the environment in which it must operate.

NVIDIA’s separate synthetic-data example turns transaction histories into fraud-classifier features. A bank would need to compare them against its own baseline and decision requirements before drawing conclusions.

A catalog listing describes an intended application, while a matching artifact lets a team examine the implementation and its dependencies. The transaction listing’s GitHub link led to a different project when checked, leaving the availability of that specific Cloudera implementation unconfirmed.

Read the full story
—
Back to the briefing
news

Microsoft profit growth and its OpenAI investment

How an investment adjustment changes Microsoft’s profit comparison.

The big change

Microsoft earns revenue from its commercial relationship with OpenAI and accounts separately for its investment. These relationships answer different questions about the company’s financial performance. Sales produce revenue; investment earnings are accounted for separately.

A prior-year investment loss lowers the base for reported growth, making the adjusted comparison slower. Both measures rise, but they give shareholders different comparisons.

Profit and cash spending describe different parts of a business. Cash paid for long-lived assets is a separate question from income reported in a period, so a profit growth rate alone cannot establish the return on a particular AI infrastructure project.

Read the full story
—
Back to the briefing