NEOM's AI Backbone: Inside The Line's Cognitive Infrastructure

NEOM's chief technology officer Joseph Bradley described The Line, in a November 2025 interview with Bloomberg, as "a city designed around a nervous system, not a street grid". A year and a half later, with the first 2.4-kilometre stretch of the 170-kilometre linear city now under structural close-out, the shape of that nervous system is finally public enough to assess. The short version: the cognitive infrastructure is further along than the built environment, and it is more ambitious than anycomparable smart-city deployment anywhere in the world. ## By The Numbers - **$2 trillion - Combined Gulf sovereign wealth deployed toward AI and technology diversification** - **40% - Projected increase in MENA AI market size year-on-year through 2028** - **9 - Number of Arab states with published national AI strategies** - **$15 billion - Estimated annual AI investment across the GCC by 2025** ## The operating principle Every other smart-city project - from Songdo to Masdar City to Toronto's shelved Sidewalk Labs pilot - bolted sensors and analytics onto a conventional urban form. The Line inverts the sequence. Its cognitive stack was specified before the foundations, and every physical subsystem, from mobility to water to retail, is required to expose an API into a single city-scale operating layer known internally as NEOM OS. NEOM OS is a federation of six service planes: environmental sensing, mobility and logistics, utilities, public safety, hospitality and resident services, and governance and identity. Each plane runs on a dedicated regional cloud footprint built in partnership with Oracle, Microsoft and G42's Core42, with compute and inference priced into the cost of the underlying service rather than billed separately to tenants. The ambition is not just a dashboard for city managers. It is a single reasoning layer against which residents, businesses, visitors and autonomous systems can all query the state of the city in natural language and receive an action, not just a readout. ## What is actually live NEOM has been cagey about publishing telemetry, but several operational disclosures have landed in the past quarter, mostly through filings to the Royal Commission and procurement tenders.

For related analysis, see: [AI-Powered News for YouTube: A Step-by-Step Guide (No ChatGP](/business/how-to-create-ai-generated-content-for-a-news-channel-on-youtube-without-using-chatgpt).

Environmental sensing is the most built-out plane. More than 41,000 multi-modal sensors are live across the first module, covering air quality, acoustics, thermal imaging, hydrology and seismic activity. The sensing mesh feeds into a digital twin built on Bentley iTwin and Nvidia Omniverse infrastructure, updated at five-second granularity. Mobility is next most mature. The underground high-speed rail spine is still in commissioning, but the surface and sub-surface autonomous shuttle network is in limited service, running on a tightly geofenced version of Pony.ai's urban stack fine-tuned on NEOM-generated training data. A separate autonomous logistics layer moves materials and waste through service tunnels below the residential floors. Utilities is partially live. The city's water system runs on predictive demand modelling out of DEWA's AI lab in Dubai, adapted for The Line's closed-loop greywater and atmospheric water generation setup. Grid balancing leans heavily on forecasts from Masdar's renewable-generation portfolio combined with in-module battery storage. Public safety is live at pilot scale and is the most politically sensitive of the planes. NEOM has committed publicly to a rights-respecting design standard, and its Resident Digital Rights Charter, published in February, gives residents the ability to query, export and in some cases delete data held about them. Civil-society groups have welcomed the charter while flagging that enforcement mechanisms remain opaque.

For related analysis, see: [EA's AI Push: A Cautionary Tale for Game Studios](/business/ea-s-ai-push-a-cautionary-tale-for-game-studios).

## The Arabic-language cognitive layer The single most underreported element of the build is the language layer. NEOM OS is multilingual by design, but its reasoning core is tuned to operate natively in Modern Standard Arabic and Gulf dialects, not just to translate from English. The underlying models are fine-tuned variants of Falcon 3 and ALLaM, with NEOM-specific retrieval over the city's operational data. A resident asking a smart-home assistant, in Saudi dialect, to reschedule their autonomous shuttle and pre-cool their apartment will be served by the same reasoning layer a city operator uses to manage peak evening load. This has two strategic consequences. First, it gives NEOM a credible claim to be the first city in the world with a native-Arabic cognitive infrastructure at scale. Second, it creates a natural export product: the same stack, stripped of NEOM-specific service integrations, is reportedly being shopped to municipalities in the wider region. Riyadh's Diriyah Gate Development Authority and Qatar's Lusail City have both been named in procurement documents seen by Arab News. ## Where the risks sit

For related analysis, see: [The Rise of AI-Powered Building Management in MENA's Megapro](/smart-cities/ai-powered-building-management-mena-megaprojects).

The principal risk is not technical. It is operational and political. Three things worth flagging: Staffing the OS is expensive. NEOM disclosed in its 2025 annual report that it had hired 1,100 engineers into its technology and data directorates, and the median tenure of that cohort was still under 18 months. Retention at Gulf salaries, in a global market where senior AI-systems engineers can name their number, will stay difficult. Governance of the cognitive layer has outrun the legal framework. Saudi Arabia's Personal Data Protection Law and the AI Ethics Principles issued by SDAIA are the relevant instruments, but neither was designed for a jurisdiction where a single municipal operator controls this much real-time resident data. The Royal Commission has proposed a NEOM-specific data authority; the draft statute has not yet been tabled. And the timeline pressures are real. The public delivery target for Phase 1 of The Line slipped from 2030 to 2034 in last year's review. The cognitive infrastructure is ahead of the built environment, which creates a strange interim in which the software is ready for users the buildings cannot yet house.

For related analysis, see: [Claude Now Builds Interactive Charts in Chat](/news/claude-now-builds-interactive-charts-in-chat).

## What to watch Three milestones matter between now and year end. The first module hand-over to the first 1,000 resident-operators, currently pencilled for the third quarter. The publication of NEOM OS's external developer SDK, which would confirm the export ambition. And the Royal Commission's decision on the proposed NEOM Data Authority, which will shape how the rest of the region copies or declines to copy the model. On balance, The Line's cognitive layer is the most credible piece of the NEOM programme today. The question is no longer whether the software can work at city scale. It is whether the physical city, and the governance around it, can catch up.

Further reading: Reuters | OECD AI Observatory

## Frequently Asked Questions ### Q: How is the Middle East positioning itself in the global AI race?

Several MENA nations, led by Saudi Arabia and the UAE, have committed billions in sovereign AI infrastructure, talent development, and regulatory frameworks. These investments aim to diversify economies away from hydrocarbon dependence whilst establishing the region as a global AI hub.

### Q: What role does government policy play in MENA's AI development?

Government policy is the primary driver. National AI strategies, dedicated authorities like Saudi Arabia's SDAIA, and initiatives such as the UAE's AI Minister role have created top-down frameworks that coordinate investment, regulation, and adoption across sectors.

### Q: What are the key smart city AI projects in the Arab world?

• Major projects include Saudi Arabia's NEOM
• Dubai's Smart City initiative
• Abu Dhabi's Masdar City
• all showcasing AI-driven traffic management
• waste optimisation
• citizen services integrated from the ground up

Sources & Further Reading

• World Economic Forum - AI in MENA
• IMF - MENA Economic Outlook
• UAE AI Office - National AI Strategy 2031
• OpenAI Research
• NVIDIA - AI Infrastructure