For years, the conversation about artificial intelligence was binary:
Are you using it or not?
At the beginning of 2026, that question is not only naïve – it is dangerous.
AI is no longer an external tool. It is inside processes, inside decisions, inside the network, inside the device… and inside people’s exhaustion.
The right question now is another one:
Who governs the intelligence that is already operating your system?
These five trends are not about promises or demos. They are about real friction, and real decisions that companies, technology platforms and operators are already taking. Decisions that make the difference between using AI… and becoming trapped by it.
1️⃣ The AI Operating Model: the advantage is no longer the model, it is the organization
Between 2024 and 2025 something curious happened:
Almost every company “adopted AI”… and yet, few achieved sustainable results.
The problem was not the technology. It was the absence of a clear operating model.
The most common mistake
Many organizations treated AI as:
- just another tool
- a plugin
- a magic layer on top of broken processes
But AI does not work like that. AI amplifies what already exists.
- If there is disorder, it amplifies disorder.
- If there is ambiguity, it amplifies ambiguity.
What the companies that are moving forward are doing differently
The most mature organizations have started to build something less glamorous, but infinitely more powerful: an AI Operating Model.
This means explicitly defining:
- which decisions an AI can take without supervision
- which ones require human validation
- which metrics define “a good decision”
- who is responsible when something fails
- how cost, quality and risk are controlled
Not in presentations. In real operating rules.
The cultural shift
The conversation stops being:
“Which model do we use?”
And becomes:
“How does our organization work when intelligence is no longer only human?”
In 2026, competitive advantage is not having the “best AI”, but the best choreography between people, processes and artificial intelligence.
Concrete signal for telcos
McKinsey’s work on “agentic” organizations shows that around 89% of companies remain in industrial structures and only about 1% operate as decentralized networks of teams and agents. Where they redesign a domain as AI-first, some banks report up to 50% less time and effort to modernize legacy systems, with small human squads supervising factories of AI agents.
For telcos this is the real lever: not more AI projects, but a few high‑impact domains (provisioning, assurance, billing, care) run by expert teams that govern fleets of agents with clear metrics, playbooks and stop‑conditions.
Source: McKinsey – “The agentic organization: Contours of the next paradigm for the AI era”
2️⃣ Apple + Gemini: when AI stops being proprietary and becomes orchestrated
For more than a decade, Apple built its identity on a clear idea: absolute control of the ecosystem.
That is why the decision to integrate Gemini, Google’s AI model, into its intelligence architecture is not a technical detail. It is a carefully calculated strategic break.
What this decision really means
Apple is not abandoning its vision. It is evolving it.
- It continues to prioritize on-device AI.
- It continues to put privacy and experience at the center.
- But it accepts an uncomfortable truth: no single company can cover the entire spectrum of intelligence alone.
The result is a new pattern:
- local AI for speed, context and privacy
- cloud AI for capacity, reasoning and scale
- a system that decides dynamically which intelligence to use, when and for what
The device stops being just “smart”. It becomes an orchestrator of intelligences.
Why this matters beyond Apple
This movement redefines:
- the role of the device
- the balance between edge and cloud
- the strategic value of the network
- and the relationship between platforms and operators
For telecommunications, this is not neutral:
More distributed AI means more intelligent traffic, more dependence on low latency and more pressure on the network as a strategic asset, not a commodity.
The future is not “my model versus yours”. It is who governs the flow of intelligence in a distributed system.
Concrete example and numbers
According to Marketing Dive, Apple and Google closed a multiyear deal where Apple is expected to pay around $1 billion per year for access to Gemini, on top of the roughly $20 billion a year that Google already pays to stay the default search engine on Apple devices.
In telecom language: value is concentrating in who owns or rents the intelligence layer. Every new on-device or cloud AI feature generates more context‑rich traffic, more sensitivity to latency and far more strategic conversations about who controls data paths and inference workloads.
Source: Marketing Dive – “Apple taps Google Gemini to power AI features in multiyear deal”
3️⃣ Network intelligence: the network stops obeying and starts deciding
For years, artificial intelligence in telecommunications was descriptive:
- dashboards
- KPIs
- alerts
- post-mortem reports
In 2026, that is no longer enough.
Modern networks — 5G SA, slicing, edge, extreme virtualization — are too complex for direct human operation.
The real shift
AI begins to:
- anticipate failures before they occur
- redistribute capacity dynamically
- optimize energy consumption in real time
- prioritize traffic by context, not just static rules
This is not marketing. It is operational survival.
The new role of the operator and the engineer
The network is no longer “constantly configured”. It is supervised, audited and governed.
Human value moves towards:
- defining the right policies
- validating automatic decisions
- intervening only when the system cannot resolve a situation
This demands new maturity: Less manual control, more systemic responsibility.
Telecom operators that understand this will stop competing only on coverage or price. They will compete on quality of operational intelligence.
Concrete signals for network operations
Bain notes that AI is becoming central across the telecom value chain, and NVIDIA reports that 97% of telcos are already implementing or assessing AI projects. The debate is no longer if AI enters the network, but how deep you allow it to act.
For NOCs this means moving from reacting to alarms to supervising predictive systems that reroute traffic, open tickets and trigger work orders before customers feel the impact. Operators that track downtime avoided, energy saved and truck rolls eliminated will have a stronger story for regulators and boards about why “network intelligence” now rivals spectrum in strategic weight.
Source: Bain & Company – “AI in Telecommunications”
4️⃣ Gamification + AI: learning like we play (and working better because of it)
Video games solved decades ago a problem that the corporate world still drags around:
How to keep people motivated while they learn complex things.
In 2026, AI allows bringing those mechanics to the heart of the company.
What is changing
The combination of AI + gamification enables:
- immediate feedback
- continuous learning
- realistic simulations
- visible skill progression
Especially in areas where traditional training is slow, expensive or inefficient:
- sales
- support
- leadership
- technology adoption
Why it works (when it works)
Because it reduces friction:
- micro-challenges instead of endless courses
- deliberate practice instead of abstract theory
- measurable progress instead of purely subjective evaluations
Poorly designed, gamification becomes noise, stress and toxic competition. Well designed, it becomes a living learning system.
Concrete use cases in corporate training
Specialized providers such as GameStrategies describe AI gamification projects where sales and support teams train in realistic, AI-driven simulators that adapt difficulty in real time and are evaluated under the Kirkpatrick model (satisfaction, learning, behavior, business results).
The promise is not just “more fun”, but higher completion, faster ramp-up and tighter alignment with KPIs like conversion rate, NPS or average handling time — especially for NOC engineers, call center agents using copilots and sales teams selling AI-powered products.
Source: GameStrategies – “How to apply AI gamification in corporate training”
5️⃣ AI fatigue: the human cost of continuous progress
We arrive at the most uncomfortable paradox of 2026.
People say:
“I am tired of AI.”
But they try to work without it… and they cannot.
What is really happening
The fatigue does not come from AI itself. It comes from:
- too many tools
- too many changes
- too many small decisions
- too little stability
We live in a state of “permanent update” that the human brain is not designed to sustain.
The organizational impact
This translates into:
- lower confidence in decisions
- frustration with pilots that lead nowhere
- silent burnout of key teams
The solution is not to slow down AI. It is to put it in order.
- fewer loose tools
- clearer flows
- more shared rules
- less eternal beta
- more stable systems
What science is already telling us
Neuroscience-based analyses like DeepFA’s “AI-Induced Cognitive Fatigue” remind us that the brain weighs only 2% of our body but consumes around 20% of its energy. Heavy AI use for coding, writing or data analysis can reduce glucose levels in the prefrontal cortex by up to 30%, and our working memory holds just 4–7 elements at a time.
For teams that spend hours with copilots and agents, that is an operational risk: more fatigue and bad decisions. Simple policies — AI‑free blocks, explicit “60-30-10” rules and protected deep‑work windows — become part of governance, not perks.
Source: DeepFAIR – “AI-Induced Cognitive Fatigue: When Your Brain Gets Tired of AI”
Closing: intelligence is already here, the question is whether we know how to live with it
AI is no longer an experiment. It is cognitive infrastructure.
And like any infrastructure, it needs:
- design
- governance
- limits
- humanity
The organizations that will win are not the ones that adopt fastest, but the ones that operate with the best judgment.
The question is not whether you use AI. It is whether your system — and your people — are ready to live with it without breaking.
📚 Recommended readings
“The executive’s AI playbook”
McKinsey’s playbook for executives on where AI is already creating value, how to prioritize use cases and what governance and operating-model changes are needed to scale it responsibly.
👉 https://www.mckinsey.com/capabilities/quantumblack/our-insights/the-executives-ai-playbook
McKinsey & Company
“Why agents are the next frontier of generative AI”
An article that explains why AI agents will sit on top of large language models, orchestrating workflows and decisions — very aligned with the “agentic” operating models we discuss in this post.
👉 https://www.mckinsey.com/capabilities/tech-and-ai/our-insights/why-agents-are-the-next-frontier-of-generative-ai
McKinsey & Company
“How telecoms can thrive in the age of generative AI”
World Economic Forum and Accenture outline how telcos can move from traditional CSPs to “techcos” by using GenAI to cut costs, grow revenue and improve customer experience across networks and operations.
👉 https://www.weforum.org/stories/2025/02/how-telecoms-can-thrive-in-age-of-generative-ai/
World Economic Forum / Accenture
✍️ Claudio from ViaMind
Dare to imagine, create, and transform.
FEATURED POSTS
FEATURED CATEGORIES
ABOUT ME
ViaMind Ecosystem
- ViaMind — The brand hub for ideas, technology and strategy.
- NeuraPRO — Practical software for managing real operations.
- Consulting — Advisory for complex AI and telecom decisions.