Presented by

Want to appear here? Talk with us

Fred FinOps is at FinOps X 2026, taking thirty-minute meetings on AI cost. No booth queue.

What to bring:

Google Cloud customers are getting hit with surprise bills in the tens of thousands of dollars after bad actors steal their API keys and run expensive AI model operations.

In one case, a CEO watched his bill jump from under $50 per month to over $10,000 in minutes as charges piled up for video generation and image creation services he had never activated.

The API Key Problem: Google previously told developers to place their Maps API keys in public-facing code so websites could display map widgets. About three years ago, Google quietly allowed those same public API keys to also access expensive Gemini AI models. Security researchers found 3,000 exposed API keys that followed the old Maps naming convention and could now be used to rack up massive AI bills. Bad actors found these and used them to generate videos and images at the customer's expense.

Spending Caps That Don't Cap: Multiple customers report setting hard budget limits at $250, only to find themselves charged $10,000 or more. Google automatically upgrades spending limits without user approval if an account has spent $1,000 total in its lifetime and is more than 30 days old. The system can bump customers from a $250 cap to a $100,000 cap based on these criteria.

The Refund Dilemma: Google initially told affected customers it found no evidence of fraud and refused refunds. Customers face a tough choice: dispute the charges with their credit card company and risk losing access to Google Cloud services their businesses depend on, or pay bills for services they never authorized.

If your organization uses Google Cloud, audit every API key in your environment right now, especially older ones created for Maps or other public-facing services.

Read More

Infrastructure is moving faster than FinOps tools built for the pre-AI era can track. AI workloads shift weekly. Kubernetes adds cost layers by the hour.

Extend Summit 2026 puts the seven companies solving this on one stage, and shows where Finout sits at the center of it.

📅​ Date: June 4th
🕗​ Time: 12:00 PM EST / 18:00 CEST
📍 Online

Subscribe to our Events

Brought by CloudZero Cloud Cost Playbook

CloudFront Premium flat-rate plans now let you pick your own usage tier, so you only pay for what you need and avoid surprise overage charges.

AWS Security Hub Extended adds 21 partner tools on one unified bill, with automatic EDP eligibility to help lower your total security spend.

AWS Organizations now allows more and larger Service Control Policies, making it easier to write detailed cost controls and governance rules at scale.

Read All AWS Updates

BigQuery reservation groups are now generally available, making idle slot sharing more predictable and helping teams get more value from reserved capacity.

Cloud Spanner now keeps change streams for 7 days by default — review your retention settings now to avoid unexpected storage cost increases.

Read All GCP Updates

FinOps Toolkit v14 adds a Copilot agent, Advisor recommendations, and a commitment discount dataset to make reservation and savings plan decisions easier and faster.

Read All Azure Updates

Industry expert Vijay Prankumar and Kiran Jain (CEO of FinOpsly) break down the accountability gap, shadow AI spend, governed automation, and the steps leaders must take to connect cloud and AI costs to real business value.

A Reddit discussion in the FinOps community asked practitioners to share their biggest challenges. The answers paint a clear picture of what keeps FinOps teams up at night.

The People Problem: Getting decision makers to take action topped the list. This shows up in the State of FinOps report as the most cited challenge year after year.

The Visibility Gap: Teams struggle to connect costs to the people and projects that create them. Finance sees the numbers but can't trace them back to decisions. Engineers make the decisions but never see the numbers. The result is a lot of finger pointing and not much progress.

The Timing Problem: FinOps works backward. You find out about a cost problem days or weeks after someone made the change that caused it. By then the code is in production and nobody wants to rewrite it.

The Framing Problem: How you talk about FinOps matters more than the data you show. Starting with "we need to cut cloud costs" makes people defensive. They hear "you've been doing your job wrong." Framing it as "let's invest cloud spend in things that matter to the business" gets people engaged. Same outcome, completely different reaction.

The Ownership Problem: Everyone agrees FinOps is important. Nobody prioritizes it until something embarrassing shows up in a quarterly review. SaaS subscriptions, cloud resources, and tool sprawl pile up because everyone assumes someone else is handling it. You can buy all the dashboards you want, but if nobody asks "why are we paying for this?" every month, nothing changes.

The thread reveals a hard truth: FinOps fails because of culture, not technology.

Read More

A global manufacturing company lost 22% of its cloud spending to waste. Dom Reid shares a story that many tech and finance leaders will recognize: different teams working in isolation, duplicated resources, and spending that no one tracked properly.

4 Areas That Matter Most: Companies that get governance right focus on these four things

A Simple Test: If someone asked you right now for a complete view of your technology spending across cloud, AI, software, and licenses, how long would it take? If the answer is anything longer than a few minutes, you have a gap in your governance.

Read More

A Microsoft AI architect has built a practical framework for tracking what AI agents actually cost your business, not just what they deliver. The framework breaks costs into four clear buckets.

The system works by connecting Azure cost data with agent runtime events and value records. Azure tags provide the metadata layer. Distribution keys provide the allocation rules. The result is a cost ledger that sits alongside the value ledger.

Now instead of saying "this agent saved five hours," you can say "this agent saved five hours, used this much in direct costs, received this share of platform costs, and has this amount we still cannot allocate." That is a conversation finance teams and technical leaders can actually use.

The framework uses standard Azure tagging like agentid, workloadid, businessprocess, and valuestream. But tags alone are not enough. You also need runtime events, clear allocation rules, and a policy for handling missing metadata.

The author released the full code and implementation guide on GitHub, including sample cost data, tagging strategies, distribution keys, and optional Azure deployment templates.

Read More

The job market is hungry for certified professionals who can prove results. Don't let your company's budget leak due to a lack of specialization.

Use code: FINOPSWEEKLY_20 to get an instant 20% discount on the most prestigious certification bundles:

If you want to be in the room where the big decisions are made, you need to be at FinOps X. We’ve secured preferred access for our community:

FinOps Legend on a Sabbatical

Connect with Professional