Welcome to the first edition of TASC AI News. Every fortnight we cut through the noise to the AI developments that move enterprise decisions — globally, across the UAE and KSA, and inside TASC. Each story gives you the development, why it matters to leaders now, and where relevant, what to do about it.
What's moving in AI worldwide
Agentic AI moves into production
The question is no longer whether AI agents work — it's who can run them at scale.
40% — of enterprise apps will use AI agents by end-2026 (Gartner)
For three years AI agents lived mostly in demos and proofs of concept. In 2026 that changed: agents are now running real workflows in production across customer service, software development, finance and operations. Gartner expects 40% of enterprise applications to embed AI agents by the end of the year, and multi-agent systems — several specialised agents coordinating on one task — have been named a top strategic technology trend.
The catch is the gap between trying and scaling. McKinsey finds 62% of organisations are experimenting with agents but only 23% have moved them into production at scale. The difference is rarely the model itself. It is the unglamorous infrastructure around it: clean, accessible data; clear guardrails on what an agent is allowed to do; and observability so teams can see what agents did and why.
For leaders, the signal is that the competitive separation now happens at the scaling stage, not the pilot stage. The organisations that invested early in data quality and governance are turning agent pilots into measurable productivity — while others are still stuck demoing.
Why it matters: The competitive gap now opens at the scaling stage. Governance and clean data — not demos — decide who wins.
Source: Google Cloud — AI agent trends 2026
The platform war shifts to governance
With frontier models everywhere, the battle has moved to the layer that governs agent memory, context and action.
23% — have scaled agents beyond pilots (McKinsey)
With capable models now available from every major provider, raw model performance is no longer the battleground. The fight has moved up a layer — to the platform that governs how agents remember context, retrieve information and act on your systems. Microsoft, Google Cloud, Databricks and Snowflake are all racing to own that layer rather than competing on the model alone.
For buyers this is a meaningful shift. The model you choose is increasingly interchangeable, but the platform that controls permissions, audit trails and data access is not. Governance maturity — can you see what every agent did, restrict what it can touch, and prove compliance — is becoming a core purchasing criterion rather than an afterthought.
The practical takeaway is to evaluate AI platforms the way you would any system with access to sensitive data and the ability to act on it: on control, observability and accountability, not benchmark scores alone.
Why it matters: Your platform's governance maturity now matters more than which model you pick. Treat it as a buying criterion.
Source: Tenfold — Top AI trends, June 2026
AI in the region: UAE & KSA
The Gulf leads the world on AI adoption
The UAE and KSA are no longer followers — they top global tables for putting AI to work.
97% — of UAE organisations are embedding AI agents — highest globally (The National)
The UAE and Saudi Arabia have moved from fast followers to front-runners on enterprise AI. Reporting this year puts the UAE at the top of the global table for embedding AI agents into workflows, while in Saudi Arabia roughly three in four organisations expect large-scale, measurable AI returns within twelve months — the highest confidence anywhere in the world. Nearly half already run AI in production across more than one use case.
This is backed by serious capital: national AI and digital-infrastructure commitments measured in the tens of billions, including major data-centre and supercomputing build-outs and partnerships with the largest global chip and cloud providers. The ambition is not simply to consume AI built elsewhere but to build and run it in-region.
The binding constraint, increasingly, is people rather than technology. The same reports that celebrate adoption also flag a widening AI skills gap — demand for AI-capable talent is outpacing local supply. Employers that pair their AI ambitions with a credible workforce plan will move faster than those treating AI as a pure technology purchase.
Why it matters: The constraint is shifting from technology to people — a widening AI skills gap.
What to do: Pair AI plans with a workforce plan: reskilling and flexible access to specialist talent.
Source: The National — UAE among top nations for AI adoption
Sovereign, physical and agentic AI
Three forces are pulling talent and capital into the region: sovereign, physical and agentic AI.
76% — of KSA firms expect large-scale AI ROI within 12 months (Deloitte)
Analysts point to three forces defining the Gulf's AI agenda in 2026. Sovereign AI is the push for national control of models, data and compute, so that critical capability and information stay in-country. Physical AI is the move of robotics and autonomous systems into logistics, energy, manufacturing and infrastructure. And agentic AI is the enterprise shift covered above — software agents doing real work.
Together these are pulling talent and capital into the region at pace, across sectors that rarely competed for the same people before. A logistics operator, a bank and a government regulator may now all be hiring for the same scarce AI and data skills at once.
For employers the implication is a tightening market for specialist talent. Building internal pipelines takes time, so the organisations that stay ahead will combine reskilling with flexible, compliant access to specialists they can deploy quickly — exactly the gap a regional workforce partner is built to fill.
Why it matters: Demand for AI-capable talent will outrun local supply across every sector at once.
What to do: Build talent pipelines and partner for compliant, flexible deployment of specialists now.
Source: Deloitte Middle East — 2026 AI trends
AI agents & your workforce
How to implement AI agents in your hiring process
AI agents are arriving in recruitment, and the natural place to start is the CV pile: the highest-volume, lowest-judgement part of hiring. Here is a practical way to introduce agentic screening without losing the human touch.
Five steps to add AI agents to your hiring workflow
- Map your current funnel. Find where recruiters lose the most time — usually first-pass CV review and scheduling — and target those stages first.
- Define the criteria explicitly. Write down the must-haves and nice-to-haves for each role so the agent scores against clear, auditable rules rather than vague 'fit'.
- Pilot on one role family. Run an AI agent (such as TASC's Hyrra.ai) alongside your recruiters on a single high-volume role, and compare its shortlist to theirs before trusting it.
- Keep a human in the loop. Let the agent rank and summarise, but have a recruiter make every shortlist and rejection — and review what it screens out, not just who it passes.
- Measure, then expand. Track time-to-shortlist and quality-of-hire. Once the agent matches recruiter quality at lower effort, extend it to more roles and stages.
Done well, AI agents take the grind out of screening and give recruiters time back for the parts of hiring that need real judgement. Start narrow, keep humans in charge, and let the results earn the next step.
Best practices for putting AI agents to work
Whether you are automating hiring, support or operations, the same principles separate AI-agent programmes that stick from the ones that stall. These apply to any agentic workflow you put into your business.
Principles to follow when you deploy any AI agent
- Start narrow. Automate one well-defined, high-volume task before anything broad or customer-facing.
- Keep a human in the loop. Let agents draft, rank and summarise; keep a person on every consequential decision.
- Be transparent. Tell the people affected — candidates, customers, staff — where and how AI is being used.
- Audit for bias and errors. Review what the agent gets wrong and who it screens out, not only its successes.
- Measure what matters. Track quality and outcomes alongside speed, so efficiency never quietly costs you fit or trust.
Treat every agent as a co-pilot for your team, not a replacement for their judgement. The companies winning with agentic AI scale it deliberately — one trusted workflow at a time.
How to implement AI: a leader's playbook
Playbook. Most AI value comes from people and process, not the model — BCG's 10-20-70 rule. Our playbook turns that into a pilot-to-production path: a short discovery and data audit against the NIST AI RMF or ISO/IEC 42001, a few governed use cases, an AI champion in every team, and a clear escalation route for when the AI is wrong.
Spotlight: AI Translation with Document Reformatting
TASC Translate. TASC Translate uses custom AI agents to translate English ⇄ Arabic and reformat your entire document for RTL — tables, charts, headers, layout, all rebuilt in one click. Translate DOCX, PDFs and PPT up to 50MB.
