Data and agentic AI: the seven trends that will shape 2026
Artificial intelligence is no longer a distant prospect or a mere supporting tool. According to Gartner’s latest predictions on data and analytics for 2026 and the coming years, AI is set to fundamentally transform this field, impacting all fronts: from talent and governance models to the rise of autonomous agents and the new semantic layers that will redefine access to data.
We are facing a genuine technological tsunami that will continue to blur the boundaries between human, machine, and organisational intelligence. In this new scenario, AI systems are no longer merely technical solutions; they have become true digital partners, capable of actively contributing to strategic decision-making and creating value for the business.
Gartner’s seven predictions serve as a roadmap for ‘surfing’ this agentic revolution, preparing for its challenges and capitalising on the opportunities that will shape the sector’s future:
1- Talent and skills in AI: the first test of survival
Talent with skills in artificial intelligence has become the first major competitive factor.
75% of recruitment processes will incorporate AI certifications
Gartner predicts that, by 2027, 75% of recruitment processes will incorporate certifications and proficiency tests in AI applied to the workplace.
It is not just a trend, it is a necessity: organisations that fail to upgrade their teams’ skills will fall behind those capable of harnessing collaboration between people and machines.
At Izertis, we are already supporting our clients on this journey, driving the adoption of AI and a data-driven culture through talent and change strategies, role-specific training pathways, AI maturity assessments and upskilling programmes that align technological ambition with the teams’ actual capabilities.
2- The end of office tools as we know whem
The emergence of generative AI and autonomous agents will pose the greatest challenge to traditional productivity tools in the last three decades. According to Gartner, this shift will reshape the market and is expected to have an economic impact of 58 billion dollars.
For organisations, the shift will be profound: there will be growing demand for solutions designed to deliver agentic experiences, featuring new interfaces, plug-ins, specific document types and formats designed from the outset to facilitate collaboration between people and AI systems.
3- Autonomous agents and the explosion of physical data
Agentic AI won’t only perform tasks: it is going to generate data on a massive scale. Gartner predicts that by 2029, AI agents will generate 10 times more data from physical environments than all digital use cases combined. This will include data on movement, spatial context and interactions between agents, which will be useful for predicting behaviour and simulating scenarios with a high degree of accuracy.
AI agents will generate 10 times more data from physical environments than from digital ones
The challenge, therefore, takes on a whole new dimension. We will need architectures capable of ingesting, processing and managing massive amounts of data in real time from IoT and cyber-physical systems.
And here, it’s not enough simply to have good models: integration and predictive capabilities must advance in tandem if we want agents to become the operational core of the AI ecosystem.
4- AI governance and self-regulated contracts
The rise of autonomous agents is also transforming the governance of AI. Gartner predicts that by 2030, half of all organisations will use AI agents to interpret governance policies and technical standards in machine-verifiable data contracts, thereby automating regulatory compliance and policy enforcement.
Half of all organisations will use AI to interpret policies and standards
But the risk is real. Gartner itself warns that 50% of failures in the deployment of AI agents will stem from inadequate governance and interoperability issues.
In the short term, unregulated decisions based on foundational models can result in financial losses and reputational damage.
The advice is clear: experiment in controlled environments, strengthen evaluation processes and thoroughly validate the context before scaling up.
5- AI-native startups and extreme efficiency
The next generation of unicorns won’t be built on the back of large workforces or multi-million-dollar funding rounds. Gartner predicts that by 2030, AI-native startups will emerge with annual recurring revenue of up to 2 million dollars per employee and valuations exceeding 1 billion dollars, driven by extreme operational efficiency rather than venture capital.
These will be highly specialised companies that solve specific problems using their own models, integrate AI into all their workflows, and offer experiences that are straightforward, easy to adopt and highly scalable.
For established companies, the message is clear: competition will no longer be measured solely by size, but by the speed at which they learn, their ability to execute and their operational efficiency.
6- Human leadership and critical semantic layers
The competitive advantage of AI will not be merely technological; it will also be profoundly human.
Gartner concludes that by 2030, 60% of organisations distinguished by AI will be led by executives who prioritise interpersonal skills, such as the ability to influence others, build coalitions and articulate a people-centred strategic vision. Human-centred leadership will be just as crucial as technical expertise.
The competitive advantage of AI will also be human
At the same time, universal semantic layers will come to be regarded as critical infrastructure, on a par with data platforms and cybersecurity.
These layers enable greater accuracy, lower costs and the coordination of multi-agent systems, preventing inconsistencies before they escalate and impact the business.
7- The risk associated with content is shifting towards AI engineering
The management of risks associated with AI-generated content is shifting to new hands. Gartner predicts that by 2028, 50% of these functions will shift from legal and cybersecurity departments to AI engineering teams, becoming directly integrated into software development cycles, data science and AI models. The aim: to design systems with embedded controls from the outset, rather than as an afterthought.
This approach will enable faster and more responsible innovation, ensuring that systems comply with ethical and regulatory boundaries, particularly in contexts where the model’s decisions need to be tailored to the user.
At Izertis, we turn these trends into tangible results: from MLOps and LLMOps to data lakehouse platforms and responsible AI frameworks, supporting organisations from strategy through to ongoing operations to accelerate their time-to-value in the agentic era.
The AI race will not be won by those who move the fastest, but by those who best combine technology, human leadership and governance