AI slowdown debate: What it means for the GCC
Regional AI leaders say the Gulf can continue accelerating adoption while strengthening governance, cybersecurity and controls around increasingly autonomous systems
18 September, 2026
TT
16
As global debate intensifies over whether the development of increasingly powerful artificial intelligence (AI) systems should be slowed, technology experts in the Gulf say the issue should not be framed simply as a choice between speed and safety.
For the UAE and Saudi Arabia, which have committed significant capital to AI infrastructure, data centres, sovereign models and wider adoption, the distinction between developing frontier models and deploying existing AI technologies could prove particularly important.
Three regional experts who spoke to Gulf Business broadly argued that the GCC can continue expanding AI adoption, while putting stronger safeguards around high-risk applications and autonomous AI agents.
Nizar Hneini, senior partner, managing director at Roland Berger Middle East, said calls to pace frontier AI development largely concern the laboratories building the world’s most advanced models rather than countries deploying existing technologies.
“The GCC should not be too worried about pacing at the frontier as it does not conflict with accelerating AI adoption. Also our region isn’t building at the frontier itself. Pacing constrains the labs racing to build the most advanced models; it doesn’t constrain how quickly the Gulf adopts what already exists.”
Hneini said GCC developers have produced leading open-source and Arabic-language models but are not operating at the same “closed frontier” as companies such as OpenAI and Anthropic, leaving room for the region to continue developing capabilities in areas where it has expertise.
Amir Grabic, co-founder and CEO of Mahala.ai, similarly argued that the debate risks presenting businesses and policymakers with a false choice.
“I’d say we’re making it a binary choice where more nuance is needed. Don’t forget that AI sits on the new world’s oil: data. In the competitive world we’re in, the acceleration should be on the foundation: getting data quality high and AI-ready. Accelerating on a solid foundation is the right way forward. “Slow down” or “speed up” is the wrong axis.”
Grabic said the Gulf’s competitive advantage is more likely to emerge from effectively applying AI to real-world business problems than from attempting to win the frontier-model race.
Dr. Charalambos Theodorou, founder and CEO of Xybern, also cautioned against reducing the discussion to a question of whether AI development should accelerate or slow.
“There is an important distinction between advancing AI capabilities and deploying those capabilities without the necessary security, governance and control infrastructure around them.”
As AI increasingly moves from generating information towards independently executing actions, Theodorou said identity, authorisation, human oversight and clearly defined operational boundaries will become more important.
Could AI guardrails attract investment?
The debate has particular significance for the UAE and Saudi Arabia because of their growing investments in computing infrastructure and AI ecosystems.
Hneini described the GCC’s position as primarily a host of computing capacity and capital rather than a frontier model builder. Demand for AI computing, particularly inference, should continue growing regardless of how quickly the most advanced models develop, he said.
“In fact, responsible AI development should strengthen investor confidence.”
Grabic agreed that regulation itself does not necessarily represent a barrier to investment.
“Order creates safety and safety attracts capital, so it really depends on what kind of guardrails. Clear, well-written ones are an asset. They signal stability and give a sense of where things are heading.”
However, he warned that vague or overly burdensome rules could restrict growth and disproportionately affect smaller companies that cannot absorb the same compliance costs as established players.
Theodorou said regulatory frameworks should similarly reflect the actual risks presented by individual AI systems.
“A model generating marketing copy does not create the same risk profile as an autonomous agent capable of interacting with financial systems or critical infrastructure.”
Clear and interoperable frameworks around security, accountability and data protection could provide companies and investors with greater certainty, while fragmented requirements could make international deployment more difficult, he added.
Extinction fears versus immediate AI risks
While much of the global discussion has focused on whether highly advanced AI could eventually pose an existential threat to humanity, the three experts highlighted more immediate risks for Gulf businesses.
Hneini said extinction scenarios remain extrapolations but argued that the risks associated with AI “misalignment” need to be addressed proportionately as models become increasingly capable.
Theodorou said there was considerable uncertainty around the probability, mechanisms and timelines of extreme AI scenarios, while businesses already face tangible challenges including cyber misuse, data exposure, misinformation, excessive permissions and unreliable autonomous actions.
“An incorrect answer from a chatbot is one category of risk. An autonomous system executing an unauthorised action against a production environment, financial system or critical operational workflow is a fundamentally different category.
“That is tangible, technically addressable and relevant now.”
Grabic also pointed to risks arising when organisations give increasingly capable systems too much autonomy without having adequate data and governance foundations in place.
“The rogue superintelligence is a spectacular danger, but the expensive agent executing damaging commands or making decisions based on bad data is the real, albeit mundane, danger.”
AI agents emerge as key GCC risk
Cybersecurity and autonomous AI agents emerged as a significant concern across the responses.
Hneini identified cybersecurity as the principal AI-related risk facing GCC businesses over the next three to five years, particularly as AI-powered cyberattacks converge with risks created by companies’ own autonomous agents.
He pointed to the region’s concentration of high-value targets — including banks, energy assets, sovereign wealth and government services — alongside its rapid adoption of AI and digital government.
“Taken together, this creates heightened cyber risk for the GCC, so companies and governments alike need to upgrade both their cybersecurity and AI governance, to ensure they are implementing agentic AI in a manner that does not compromise the systems these agents can access.”
Theodorou said the next major transition for enterprise security would be AI moving from systems that advise people to systems that can act independently.
Once AI can select tools, access infrastructure, initiate workflows and execute operational decisions, he said, the fundamental question becomes how much authority organisations are prepared to give it.
“The challenge is not to prevent AI systems from becoming autonomous. It is to ensure that autonomy never becomes equivalent to unlimited authority.”
For Gulf businesses, the experts’ responses suggest the emerging AI debate may therefore be less about whether adoption itself should slow and more about whether governance, cybersecurity, data infrastructure and human oversight can keep pace with increasingly powerful systems.
