What happened in the world of data in 2025?

Reflecting on the Data Journey in 2025

Throughout the year as our content explored the data journey in detail, there was always one theme running alongside it all, constantly bidding for our attention, and that was AI. These were the two letters on everyone’s lips in January and they’re still with us now, as we end the year, trying to assess the impact it has had on our data projects and in our day-to-day lives. Whatever you think about the wider political and social context, did the business advances roll out as predicted?

What better way to find out than by asking AI itself. So before we get into PTR’s analysis of the AI year and our very human predictions for what lies ahead, here’s what the various AI engines tell us mattered in AI in 2025:

Gemini – What mattered in the world of Data in 2025:

Key points from Gemini:

• The AI Native Shift – embedding AI into data infrastructure

• Architectural Evolution

  • Real-Time Everything – instant decision making

  • Edge Computing and AI – minimal latency with data processed closer to source

  • Modernising Architectures - Data Lakehouse

• Governance and Democratisation – focus on security and compliance

Chat GPT – What mattered in the world of Data in 2025:

Key points from Chat GPT:

• AI Became the Default Layer of the Data Stack - Productivity explosion - data teams spent dramatically less time on grunt work

• The Rise of the “Semantic Layer Everything” Era - Data consistency improved dramatically, and AI made use of these layers to give reliable answers

• The Explosion of Real-Time + Streaming Analytics - Companies monitored customer experience, fraud, ops, and supply chain faster than ever

• GenAI + Data = Operational AI Systems - Data teams became AI enablement teams

• The Corporate Data Lakehouse Won - Better cost control and flexibility; fewer data copies

• Privacy, Data Residency, and AI Regulation Became Central - Privacy engineering became a core skill for data teams

• Data Quality & Observability Tools Matured - Faster root-cause analysis and lower cloud bills

• The Skills That Mattered Most in 2025 - Less SQL monkey work; more strategic roles

Microsoft Copilot – What mattered in the world of Data in 2025:

Key Points from Microsoft Copilot:

• AI Everywhere – AI drive analytics and copilot-style assistants

• Real-Time Data Processing -streaming analytics and instant insights

• Data Governance & Trust – ethics, compliance and governance

• Automation & Efficiency – freed up human talent for higher-value strategic tasks

• Synthetic Data (artificially generated by algorithms) & Augmented Data (transformed real data to generate variations)

• Data Democratization – self service analytics

• Sustainability & Social Impact – demand for accessible and trustworthy insights

Best & Worst for Data in 2025?

So the award for innovation and progress in data in 2025 goes to……drum roll… AI!

The industry agrees, - the integration of AI and Machine Learning into day to day business is THE significant data development for 2025. It was significant in a good way in terms of all the benefits it brought, but it was significant in a bad way too – let’s not sugar coat it.

The worst because when AI is not strategically planned, back doors get left open, AI predictions are assumed to be watertight and bad decisions get made. You may even get data theft, loss of confidence, lost revenue and more.

Let’s break down the pros and cons.

AI positives:

• AI makes us faster

• AI makes us more efficient

• AI keeps us better informed

• AI reveals insights that were previously hidden

• AI automates tedious and time-consuming tasks which improves our working lives

• AI helps people do a better job and makes for a happier team

AI negatives:

• When AI is used for malicious and criminal activity it impacts a lot of people and businesses.

  • Loss of service

  • Data theft

  • Financial theft

  • Scamming

  • Deepfake

• When AI drives actions based on bad data or bad algorithms disastrous times can follow for businesses.

  • Skewing results through poor models learning from biased data leading to poor decisions

  • Damage to reputation through misrepresentation

  • Loss of revenue due to poor decisions

• Poorly secured and governed data can lead to exposure of confidential data leading to legal action, loss of reputation and confidence.

Lean into the positives but be prepared

Many businesses have learned the hard way that data governance, data security and data quality have to rank top of the task list before adopting AI processes.

Before launching into an AI implementation do make sure you have strong AI Data Governance and AI Ethical Deployment Polices and Procedures in place.

• Look at your AI process from end to end for storage, movement, processing, analysis, archiving and deleting.

• Consider legal and internal requirements concerning your AI data

• Consider the full social impact of your AI implementations such as fairness, bias, transparency, authenticity, accountability, human supervision and oversight, human rights, environmental impact

How are we using AI right now?

There has been lots of hype about AI throughout 2025 and businesses and individuals have moved from dabbling with an AI chat bot to bringing AI into more of their daily tasks and business operations.

Big scale AI transformations are not happening on as wide a scale as some might have liked or indeed predicted, but certainly more businesses are trying it out compared to the start of 2025.

Big scale AI transformations require a lot of planning, governance, procedures and policies, but there are many examples of where AI on a smaller scale is improving efficiency, productivity and job satisfaction. Let’s break it down;

• People are introducing Microsoft Copilot, Chat GPT and Gemini into their day to day lives

  • Asking for advice

  • Summarizing lengthy reports and emails

  • Asking for dashboard insights that may not be directly addressed by the dashboard design

  • Auto filling fields from web forms in apps such as Dynamics

• AI search engines have gone live such as Google AI Mode and Bing AI Mode

  • Concise summaries of search content to make it easier to find relevant web content

• Business Process Automation

  • AI agents are being used to automate business workflows to improve process efficiency, quality of data, reduction of human introduced errors.

• AI has started to move out of the IT department and into business functions.

• AI data governance and ethical deployment policies have come into play as businesses think about how their data might be exposed with the rapid growth of AI adoption.

Business are introducing Microsoft Copilot, Chat GPT and Gemini into their day to day lives

This is where the most people have started dabbling in AI and it certainly can make day to day tasks a great deal easier. For an overview of what you could be missing, have a look at this article on Making Life Easier with Chat GPT, Copilot and Gemini.

Business Process Automation

AI is being put to use automating a great many aspects of business life. The use cases for automation are limitless but take a look at this article for just 10 Ways Automation Can Accelerate Business Growth with Microsoft Power Platform

What Else Happened in 2025?

The main evolutions around Data in 2025 were either linked to AI or Automation in some way, whether it be via infrastructure, services, or tools.

With all the advances in AI and the growing interest in real-time Intelligence, instant insights, and process automation, the main players in the cloud native platform world, (Amazon Web Services AWS, Microsoft Azure and Google Cloud Platform) have all been focussing on these key areas:

• AI integration

• Improvements to core infrastructure for higher performance

Amazon Web Services (AWS)

The major focus for AWS was Agentic AI, a service that can act on its own to carry out tasks with minimal human intervention.

Google Cloud Platform

A major focus for Google Cloud Platform was also Agentic AI

Microsoft Azure Platform

A major focus for the Azure platform was also Agentic AI and the expansion of Copilot.

Microsoft Fabric is a large part of the Microsoft Azure data strategy and the Fabric platform continued to evolve throughout 2025.

Power BI also saw some great additions to its feature set.

Microsoft Fabric Evolution

In December 2024 Fabric reached its first birthday and in our December 2024 newsletter we talked about some of the new features that had been introduced in Preview to set Fabric on the way to growing into a full data platform solution.

Key releases throughout 2025 that have continued Fabric’s journey to offering a one stop fully integrated data platform are:

• January to March 2025 – Copilot availability

• April to June 2025 – Realtime analytics pipelines

• July to September 2025 – Adaptive throttling, Multi-cloud compatibility

• November 2025 – General availability of SQL Database and Cosmos DB, Mirroring support for SQL Server, Cosmos DB and PostgreSQL, Copilot chat tools, Realtime data exploration

• December 2025 - Analytics-ready data via OneLake shortcuts

Power BI Evolution

• Q3 2025 - Copilot for DAX query generation

• Q3 2025 - Table auto-size columns

What’s coming in 2026?

We have picked out a few of our favourite Microsoft Fabric features on the radar for early 2026.

Microsoft Fabric

Fabric Releases Announced:

• Fabric Materialized Lake Views

• Version history for semantic models

• Lakehouse Streaming Tables in the Real-Time hub

• Fabric IQ – currently in preview and expected ton be in General Availability Q4 2026.

Power BI

Power BI Releases Announced:

• Visualisation Calculations

• Support for multiple calendars such as 445 calendars

• Table and Matrix Fixed Width Columns

• Copilot expansion – analysis mode, author feedback

• Developer Mode – New PBIR format supporting team development

Real-Time Intelligence

Real-Time Intelligence was added into Microsoft Fabric in November 2024, but as with all things new it takes a while for adoption to gather pace.

But there has been such a lot of interest from our clients in this area that we believe 2026 is set to see the adoption rate of real-time intelligence pick up rapidly. Microsoft have some new features planned in this area for 2026.

For those who have identified a requirement for real-time BI, 2026 is the year to start exploring. As our consultant Brett Valentine-Dunn explains;

The transition to Real-Time Analytics is perhaps less of a technological shift and more of a mindset shift. We already rely on live information in every aspect of our lives - from ordering dinner to tracking our commute. It is time for our business operations to catch up to this expectation.

We will be exploring the world of Real-Time Intelligence within the Microsoft Fabric world in future newsletters, but here is Brett’s excellent taster blog and TFL/London Underground case study on What Realtime BI is all about - The transition to Real-Time Analytics

In Summary

So, where will 2026 will take us?

At the highest level the answer is undoubtedly more AI. Although some organisations have already embraced AI, 2026 is going to see many more businesses adopting it in some way.

Governance, Compliance, Trust & Authenticity are going to feature highly on the priority lists of senior management to ensure successful implementations of AI solutions.

Business Automation is going to grow with more businesses adopting the Agentic AI approach to getting tasks done in an efficient and less error-prone manner.

AI will be integrated with Business Operations and BI on a more widescale basis in 2026 as confidence grows in the platforms, tools and services, combined with a better understanding of data governance and trust requirements, and internal policies to support them.

Real-Time Intelligence will likely take off in 2026 with higher demand for instant insights, with cloud platforms offering high performance frameworks for minimal latency, instant response performance.

Business may start to opt for Data Lakehouses rather than the traditional Data Warehouses to support decentralised data sources, cost effectiveness, scalability and governance.

Regular business users will ask their managers to provide better training in already available AI tools like Co-pilot so they can level up to the same degree of efficiency as their competitors and peers.

We’d love to know what you’ve noticed about the year gone by and what pre-occupies you for the future. And if you need to lean into any of these skills with training or consultancy you know where to come.

Best Wishes,

Mandy

Mandy.doward@ptr.co.uk