The architects: Building the AI foundation

According to Gartner, architects are focused on “creating AI-ready foundations that enable speed, security and scalability.” Practically speaking, this looks like:

• Using AI tools to easily and quickly build and customize software

• Smaller teams doing what once required entire departments

• Scaling infrastructure to support more ambitious models

But building faster, better, and more secure AI systems also creates more opportunities for things to go wrong. With lower barriers to entry, the need for stronger training and evaluation will increase — keeping humans essential in shaping new systems. Gartner highlights three trends that fall into this category. 

AI-native development platforms 

What is it?
Tools that use AI to help people build software faster, often by turning plain-language instructions into working code or workflows.

Why does it matter?
These platforms lower the barrier to building AI-powered tools, which means more models, more applications, and faster experimentation across industries.

Where’s the human angle?
When AI is easier to build, the quality of its training matters even more. Human feedback can help these rapidly built systems behave as intended, not just as designed.

AI supercomputing platforms 

What is it?
High-powered computing environments that make it possible to train and run larger, more advanced AI models.

Why does it matter?
As models grow more complex, they need more computing power to learn, adapt, and perform reliably at scale.

Where’s the human angle?
Bigger models don’t automatically mean better ones. AI trainers will play a critical role in shaping how these powerful systems respond, reason, and communicate.

Confidential computing 

What is it?
Technology that protects data while it’s actively being used by AI systems, not just when it’s stored or transferred.

Why does it matter?
It allows organizations to use sensitive or regulated data more safely, opening the door to AI in areas where trust and privacy are non-negotiable.

What’s the human angle?
As AI systems handle more sensitive information, AI trainers will become essential in ensuring outputs are responsible, fair, and aligned with real-world expectations.

The synthesists: Helping AI work together

Generalized AI is starting to give way to more specialized systems. According to Gartner, synthesists focus on “orchestrating diverse technologies to create adaptive, intelligent ecosystems.” Practically speaking, this looks like:

• Multiple AI systems working together, each focused on a specific task

• Models trained for specific domains rather than one-size-fits-all use

• AI being embedded more deeply into real-world workflows and tools
  
  
  

Gartner predicts that at least 60% of enterprise GenAI models will be domain-specific by 2028. While this shift has high potential, small errors can also ripple across systems and threaten consistency. That’s where careful training, evaluation, and human judgment will be critical. Gartner highlights three trends in this category.

Multiagent systems

What is it?
Instead of one AI doing everything, multiagent systems use several specialized AIs that work together, each handling a different part of a task.

Why does it matter?
This approach makes AI better at complex, multi-step work, but it also introduces more moving parts that need to stay aligned.

Where’s the human angle?
When multiple AIs collaborate, consistency matters. Human trainers can help shape how agents interact, hand off tasks, and stay on track, especially in complex cases when systems don’t agree.

Domain-specific language models

What is it?

AI models trained on focused industry- or task-specific data rather than broad, general-purpose information.

Why does it matter?
Specialized models are often more accurate and useful in real-world settings, but only when they truly understand the context they’re built for.

Where’s the human angle?
Domain-specific AI depends on high-quality human input. AI trainers with subject-matter expertise can help models learn what’s relevant, what’s risky, and what “good” actually looks like.

Physical AI 

What is it?
AI systems that interact with the physical world, like robots, sensors, and automated machines that can observe, decide, and act.

Why does it matter?
As AI moves out of the screen and into real environments, mistakes can have real-world consequences, not just bad outputs.

Where’s the human angle?
Even the most advanced physical AI needs human guidance. Thoughtful training and evaluation can help these systems make decisions that are safe, reliable, and aligned with human expectations.

The vanguards: Keeping AI trustworthy

According to Gartner, Vanguards are the guardians of safety, focused on “proactive security, transparent governance and digital integrity.” Practically speaking, this looks like:

• Creating AI systems that can better protect themselves from misuse or attacks

• Establishing clear ways to understand where AI outputs come from

• Shaping stronger safeguards around how AI tools are built and deployed
  
  

But trust doesn’t come from technology alone, it’s built through oversight, transparency, and accountability. Gartner highlights four trends that fall into this category.

Preemptive cybersecurity

What is it?

Proactive security approaches that use AI to anticipate and stop threats before damage is done instead of reacting when something goes wrong.

Why does it matter?

AI systems introduce new kinds of risks, and traditional security tools aren’t always equipped to handle them fast enough.

Where’s the human angle?

Human judgment helps define what “safe” looks like. AI trainers can play a role in identifying risky behavior, edge cases, and unintended outcomes that automated defenses might miss.

Digital provenance 

What is it?

Ways to track where data, software, and AI-generated content come from, and whether they’ve been altered.

Why does it matter?

As AI-generated content becomes more common, it’s harder to tell what’s real, reliable, or trustworthy without clear signals of origin.

Where’s the human angle?

Provenance starts with responsible training. Human input helps establish credibility, context, and quality long before outputs are labeled or traced.

AI security platforms 

What is it?

Tools designed specifically to protect AI systems, including how models are prompted, how they act, and how data flows through them.

Why does it matter?

AI introduces new failure points that traditional security tools weren’t designed to handle.

Where’s the human angle?

Humans can help surface weaknesses early by testing boundaries, spotting unexpected behavior, and helping systems learn what not to do.

Geopatriation 

What is it?

The shift toward running AI systems in local or region-specific environments to reduce regulatory and geopolitical risk.

Why does it matter?

Where AI systems run (and where data lives) increasingly affects compliance, trust, and long-term stability.

Where’s the human angle?
Training and evaluation will need to reflect regional norms, rules, and expectations. Human input can help models adapt responsibly across contexts.

What Gartner’s trends don’t cover (hint: it’s humans)

Gartner’s trend overview highlights how organizations will scale, connect, and secure AI systems. But focusing only on systems leaves out a critical piece of the puzzle: the people who make AI more reliable, useful, and human-centered.

As we explored in our own look at the biggest AI trends for 2026, the human contribution is where value really gets unlocked. For example: 

• AI agents are no longer just prototypes — they’re increasingly acting like teammates that need careful coaching and evaluation to handle complex tasks well. 

• Embedded AI tools are becoming part of daily workflows, but their usefulness depends on how well they’re trained to understand context and user intent. 

• Privacy-first AI (an emerging priority for 2026) only fulfills its promise when humans help design systems that protect data and deliver value without compromising trust.

So while Gartner maps the technological transformations emerging throughout the next year, the true differentiator will be the people behind the AI.