GCC AI News — 6 Oct 2026

Ai Everything Abu Dhabi 2026 opened October 6 at the ADNEC Centre under the theme 'The New Intelligence Order from Abu Dhabi, the AI Capital of the World,' running October 6-7. Organised by inD (the team behind GITEX) and hosted by Abu Dhabi's Department of Government Enablement with Solutions+, a Mubadala company, as global partner, the two-day expo and summit brings together more than 350 AI-first enterprises and startups, from advanced humanoids to autonomous agents, alongside global leaders, investors and policymakers. The event showcases sovereign AI, agentic systems and AI-native economies, positioning Abu Dhabi as a hub where real-world deployments, not just prototypes, are put on show.

Hosting the world's largest AI-first gathering signals the UAE's shift from AI adopter to global convener. It gives Gulf startups and government teams a home-turf stage to attract investment and partnerships, and showcases the region's ambition to shape where AI goes next.

A regional founder or GovTech team can use Ai Everything as a direct line to investors and enterprise buyers, demonstrating a working deployment on home turf rather than chasing meetings abroad.

Dubai Electricity and Water Authority (DEWA) said on October 5 that it has enhanced 16 of its 21 services, 76% of its portfolio, with agentic AI in under 90 days, with the remaining five due within two months. Customers can now apply to activate electricity and water supply in about two minutes, down from eight, with the redesigned service requiring just three inputs instead of ten and three steps instead of nine, a 75% cut in completion time on roughly 210,000 transactions a year. DEWA has also moved technical-notification handling to agentic AI that can receive, classify, verify and route requests autonomously around the clock, work that previously relied on 16 coordinators. The push follows directives to deploy agentic AI across Dubai government services.

This is agentic AI in live public infrastructure, not a pilot, showing how a government utility can collapse multi-step bureaucracy into minutes. It is a concrete, citizen-facing proof that the UAE's agentic-government mandate is reaching everyday services.

A service organisation can copy DEWA's pattern: pick a high-volume transaction, strip the inputs and steps to the minimum, and let an agent handle classification and routing so staff focus on exceptions, not queues.

NVIDIA announced on October 3 that it used the Saudi Audio Dataset for Arabic (SADA), launched by the Saudi Data and AI Authority (SDAIA) with the Saudi Broadcasting Authority, to train its Nemotron 3.5 ASR multilingual speech-recognition model, nearly halving comprehension errors for Saudi dialects. The model, which supports real-time transcription across about 40 languages and dialects, initially misread roughly 55 of every 100 words in Saudi dialects; after training on 133.7 hours of Najdi and Hijazi audio, the word error rate fell to around 30%, with character-level errors dropping from 31.6% to 12.2%, plus gains in Modern Standard Arabic and English. Training took just four and a half hours on two GPUs.

This shows the strategic value of sovereign data: a Saudi national dataset measurably improved a global frontier model's grasp of Gulf Arabic. It is a template for how the region can make world-class AI actually work for Arabic speakers, instead of waiting for Western labs to catch up.

A team building Arabic voice products, from call-centre transcription to voice assistants, can now reach for dialect-tuned ASR like Nemotron 3.5, getting far lower error rates on Najdi and Hijazi speech than generic multilingual models deliver.

Dubai's Roads and Transport Authority (RTA) launched on October 5 an AI-powered system to manage and operate its bus network more efficiently across varying conditions, according to Dubai Media Office. The system monitors more than 1,100 buses and supports 26 operational scenarios, strengthening the readiness of the public transport fleet. RTA said it can cut decision-making and bus-distribution time from the previous 30 to 60 minutes to under one minute by using real-time data, machine learning and advanced algorithms to identify and route the most suitable buses. The rollout is part of Dubai's broader effort to make AI a practical operating layer across live public services.

Public transport is where AI quietly improves daily life: fewer gaps in service, faster response to disruptions, shorter waits. Compressing an hour of manual dispatch decisions into under a minute means a more reliable network for the hundreds of thousands who ride Dubai's buses.

A transit or fleet operator can apply the same approach, real-time data plus ML to match supply to demand, to cut dispatch delays and keep service steady when conditions change, without adding vehicles.

King Abdullah University of Science and Technology (KAUST) has developed a generative-AI foundation model for synthetic geology, designed to reduce uncertainty in mineral exploration across the Arabian Shield, as reported September 29. Mineral exploration has always relied on sparse, indirect data, boreholes and surface samples that leave big gaps in the picture underground, and the model aims to deepen understanding of the geological formations beneath the Kingdom. Saudi national mining company Ma'aden was impressed enough by the prototype's potential to sign a 30-month research-and-development contract to test and deploy the model on real Saudi geological data, moving it from lab prototype toward field use.

Mining is central to Saudi Arabia's economic diversification, and exploration is slow and expensive. An AI model that fills the gaps between sparse data points could mean faster, cheaper, better-targeted exploration, turning the Kingdom's push into mining into a tech-enabled advantage.

A mining or geoscience team can apply a geological foundation model to its own survey data to prioritise drilling targets, cutting the number of expensive boreholes needed before committing to a site.

Inception42, the Abu Dhabi AI products company in the G42 group, has launched Seraj, an Arabic-first enterprise AI model built jointly with Microsoft and available through Core42's sovereign AI platform Compass. Rather than training from scratch, Seraj is built on OpenAI's GPT-4.1 and enhanced through targeted mid-training on curated Arabic datasets covering linguistics, cultural knowledge, safety scenarios and domain-specific enterprise content. The approach lets it keep GPT-4.1's reasoning and multilingual strength while shoring up the Arabic gaps that have slowed Gulf enterprise adoption. Target sectors include government services, legal analysis, Islamic studies, financial services, education and media, where Arabic fluency and cultural context are critical.

Most frontier models were built in English and retrofitted for Arabic, with mixed results in dialect, culture and safety. A model that pairs GPT-4.1's reasoning with deep Arabic tuning, delivered on sovereign infrastructure, makes high-stakes Arabic deployments viable for Gulf governments and enterprises.

A Gulf bank, law firm or ministry can deploy Seraj via Compass for document understanding, summarisation and bilingual workflows, getting enterprise-grade Arabic accuracy on sovereign infrastructure rather than routing sensitive data through a generic model.

Bahrain's National Cyber Security Centre (NCSC) launched 'Ask Amna,' the Kingdom's first AI-powered digital assistant for cybersecurity awareness, unveiled at Marassi Galleria with NCSC CEO Shaikh Salman bin Mohammed Al Khalifa. The assistant lets residents ask cybersecurity questions, learn about digital risks and safe practices, and check whether a link is trustworthy before opening it. Ask Amna is part of the 'Bu Rashid Family' initiative, described as Bahrain's first AI-powered digital family for cyber awareness, launched under the slogan 'Cyber Awareness, A Way of Life,' which uses relatable characters and everyday situations to simplify security concepts for different age groups. The service is live on the NCSC's official website.

Most cyber-awareness campaigns are leaflets people ignore; an always-available AI assistant that vets links and answers questions in plain language meets citizens where real scams happen. It is a practical, people-first use of AI to protect ordinary residents, not just institutions.

A resident unsure about a suspicious SMS or payment link can ask an assistant like Amna to check it before tapping, turning abstract security advice into a quick, concrete safety step anyone can take.

Omantel's technology subsidiary Otech and Omani AI startup DeepAstra have launched the DeepAstra API, a locally hosted sovereign AI platform that gives businesses access to multiple AI models through a single interface. Users connect once, test different models and pay by usage, with local hosting designed to meet data-security, governance and compliance needs. The platform is aimed at Omani institutions, developers and SMEs that want to use AI without building their own computing infrastructure or integrating each model separately. DeepAstra, founded in 2025 and part of the NVIDIA Inception Program, was backed by the Oman Future Fund via Omantel Ventures, and has since become Chinese lab Z.AI's exclusive sovereign AI partner in Oman, with a planned Oman Sovereign AI Development Centre focused on Arabic AI.

Building AI infrastructure from scratch is the single biggest barrier stopping Omani SMEs from moving past experimentation. A pay-as-you-go, locally hosted gateway removes both the cost and the data-sovereignty concern in one step, helping smaller firms turn AI from a pilot into a product.

An Omani startup or SME can connect once to the DeepAstra API, compare several models on its own use case, and pay only for what it uses, keeping data in-country while avoiding a large upfront infrastructure investment.

Sidra Medicine, a member of Qatar Foundation, has integrated an AI-powered tool into its Pediatric Emergency Department to help physicians interpret children's X-rays, identify potentially urgent cases faster, reduce waiting times and strengthen collaboration between emergency doctors and radiologists. Hospital officials said all AI tools undergo validation and testing by specialist radiologists before clinical use and are continuously monitored, with AI findings routinely checked against radiologists' reports. The original diagnostic X-ray is never altered: AI analysis is shown on a separate copy alongside the original so physicians can independently review both. Sidra has also introduced AI tools in its radiology and imaging services and is working with other clinical teams on further initiatives.

Pediatric emergency care is time-critical and anxiety-heavy for families. AI that flags urgent cases faster, while keeping a radiologist accountable for every read, is a model for responsible medical AI, speeding care without removing the human judgment parents rely on.

A hospital emergency or imaging department can adopt Sidra's safeguard design, AI findings shown on a separate copy with mandatory specialist validation, to gain speed on urgent cases while keeping a qualified clinician responsible for every diagnosis.

ZainTECH, the digital-solutions arm of Zain Group, and Microsoft co-hosted 'AI-Ready Kuwait,' gathering senior policymakers and digital-transformation leaders to advance Kuwait's AI ambitions under Vision 2035. The event marked a milestone ahead of the launch of Kuwait's first AI-powered Microsoft Azure region, a development intended to reinforce data sovereignty and public-sector resilience. Organisers showcased how AI and cloud solutions deployed locally by ZainTECH with Microsoft can modernise government services and scale secure, mission-critical operations across sectors including healthcare, education, emergency response and citizen engagement. ZainTECH's leadership framed the summit as about execution, turning digital ambition into deployed services rather than strategy documents.

A local, sovereign cloud region is the groundwork that lets a country run AI on its own citizens' data safely and at scale. For Kuwait, it is the practical step that moves Vision 2035's AI goals from policy toward faster, more resilient public services people actually use.

A Kuwaiti government body or enterprise can plan to migrate sensitive workloads to a local, sovereign Azure region, gaining AI and cloud capability while keeping data in-country to meet residency and resilience requirements.