The UAE government has introduced a new benchmark to measure how well artificial intelligence systems understand local culture, and Google Gemini has come out on top in its first edition. The “AI in the Ring” Index, developed by the UAE’s Artificial Intelligence, Digital Economy, and Remote Work Applications Office, is described as the world’s first assessment focused specifically on how accurately AI models reflect Emirati identity, dialects, traditions, and national values.
In this inaugural evaluation, Gemini was assessed alongside 10 other major language models using more than 400 culture-focused questions across seven dimensions. These questions generated a total of 5,200 responses, which were then reviewed by a committee of Emirati experts. Their task was to judge which models showed the strongest cultural understanding in a UAE context, rather than simply measuring technical fluency or general accuracy.
According to the results, Google Gemini 2.5 Pro ranked first, followed by several other widely used systems. The top five models in the AI in the Ring Index were listed as: ChatGPT (OpenAI GPT-4o), ChatGPT (OpenAI o1), Cohere, and Grok, alongside Gemini. The ranking highlights that multiple global AI platforms are now being tested not only for performance, but for how well they align with local cultural norms and values.
His Excellency Omar Sultan Al Olama, Minister of State for Artificial Intelligence, Digital Economy, and Remote Work Applications, linked the initiative directly to the UAE’s priorities around national identity. He argued that as digital transformation accelerates, AI systems should reinforce rather than dilute the country’s values, and that innovation is more sustainable when it is rooted in a nation’s cultural foundations. In this context, the AI in the Ring Index is as much about governance and standards as it is about technology.
The push for culturally grounded AI is also set against a wider structural challenge: Arabic remains underrepresented online, with less than 5% of global digital content available in the language. At the same time, nearly half of Generation Z is reported to rely on AI as a primary source of information. That combination raises the stakes for countries like the UAE, which want to ensure that the data shaping these models includes accurate, relevant, and locally informed perspectives.
By evaluating how well systems like Google Gemini perform as culturally accurate AI for Arabs, the UAE is signalling that future AI regulation and adoption in the region may be tied to cultural competence as much as to technical capabilities or commercial reach. For developers, that could mean more partnerships with local institutions, more nuanced training data, and closer scrutiny of how models respond to questions about customs, social norms, and national narratives.
The AI in the Ring Index is still at an early stage, but it offers a template other countries may look at as they consider how to protect cultural heritage in an era when large language models are shaping everything from education to entertainment. For the UAE, the benchmark is also a way to position itself as an active shaper of AI standards rather than a passive consumer of imported technologies. As more details from the full report are reviewed, the discussion is likely to shift from rankings alone to how these models can be improved and audited over time to remain aligned with local expectations.
