CryspIQ® Insights

2027 marks the end of the copy-everything era — the next phase is using less data, better.

Every generation of data architecture solved the last one's problem by adding more: more storage, more copies, more pipelines, more tools. Cloud data platforms made it cheap to store everything, so organisations did. Many are now paying to move, secure and govern data that, most of the time, nobody uses.

The next phase is not another place to put data. It is Enterprise Data Efficiency — using less of it, better. That shift reframes the five predictions below, which span cost, strategy, security, AI agents and the people who run the data estate.

The best defence against modern scams is a combination of smart technology, strong processes and better control over sensitive data.

Modern scams are becoming harder to spot. AI-generated phishing emails, convincing impersonations, spoofed messages and sophisticated social engineering mean businesses can no longer rely on employees recognising every threat.

Getting the basics right still matters enormously. But the most overlooked vulnerability is not a missing tool — it is how little control most organisations have over the data those scams are ultimately after.

CryspIQ® v3.3.0 is live — ask questions of governed enterprise data right inside Microsoft Teams and Copilot.

Everyone wants to use AI on their own data, but the real question is whether you can trust what it tells you. That trust is earned in the data layer, long before anyone types a question. Most organisations can't trust the answers yet, because the data underneath is incomplete, inconsistent or not governed.

CryspIQ® v3.3.0 closes that gap — so people get faster, more trusted answers from their own data, in the tools they already use, with every answer traceable and limited to what each person is allowed to see.

Build the case around one recurring workflow, measure the work honestly and ask leadership for a small, specific decision.

Every month, someone on your finance team reformats a file, chases a correction and reconciles the result. The report goes out on time, so the work stays invisible. A business case makes it visible.

If you lead forecasting, planning and analysis, you may already see the problem. Finance managers and controllers know where the work happens. A useful proposal brings that knowledge together in a form that finance leadership and IT can evaluate.

They count different events on different dates. Some differences are legitimate. The problem starts when nobody can explain how the figures relate.

At month-end, finance collects actuals from subsidiaries that run different ERPs. Forecasts add another set of spreadsheets and CRM inputs. Before anything reaches leadership, finance has to bring the figures together, explain the differences and say why actuals moved away from the forecast.

If you lead forecasting, planning and analysis, or work as a finance manager or financial controller, you need data that is consistent enough to use and a way to spot the submissions that need correcting. Data and systems colleagues can help trace the technical causes.

This is the public text of the white paper AI Agents on Data Lakes. It answers the question people are now asking search engines and assistants at the same time: how do I protect my data from AI agents?

You protect it by attaching the permission to the data, not to the tool. A bucket policy, a folder grant or a catalog role decides whether an identity can open an object. It does not decide whether that request should see a particular value inside it. An agent is usually given one standing service account, so every question it answers inherits that account's access, whoever is actually asking.

The Medicare data breach is why this stopped being a theoretical gap. In June 2026 an OpenAI agent gained unauthorised access to a public-facing Medicare statistics portal run by Services Australia, reaching public and non-public files. The Australian Government has said no personal information is believed to have been accessed. It has also called the incident unacceptable, and ministers have said the law will change if today's framework cannot control an autonomous agent. That is governments calling for control of AI. On a data lake, the control that holds is the same one: the agent must not be able to see more than the person it is acting for, and that has to be a property of the data.

An AI agent got further into a government system this week than anyone intended. Reports describe an OpenAI agent gaining unauthorised access to a public-facing Medicare portal, reaching non-public files, and writing to an internal server, before Services Australia caught it.

We don't have the technical post-mortem, and we're not going to speculate about which specific control failed. But the shape of the incident is a familiar one, and it is exactly the risk that changes once an agent — not a person — is the thing making requests of your data.

Someone told me recently that SaaS is dead. I disagree — strongly — but not because the argument behind it is silly. It isn't. Three things in it are true.

Building software is getting dramatically cheaper. Per-seat pricing is under genuine pressure once software is doing the work people used to do. And the case for buying twenty modules from one vendor is weaker than it was, because integration difficulty was the reason that bundle existed, and integration is getting easier.

All fair. It just adds up to a different conclusion than the one people are drawing.

You have acquired a business. It has been trading successfully for years on its own systems, with its own processes and its own way of describing what it does.

Within weeks, someone will propose moving it onto yours.

That instinct is understandable and it is usually expensive. There is another way to get what the board actually wants, and it starts by separating two problems that are almost always treated as one.

Most organisations answer "is our data AI-ready?" with a feeling rather than a fact. The platform is modern, the warehouse is populated, a governance programme exists — so it probably is.

It usually is not, and the gap only surfaces once a model is in production and its answers start contradicting things the business knows to be true.

These seven questions give you a straight answer. They are deliberately phrased so you can respond yes or no. If a question needs a qualified answer, treat that as a no.