IMF Warns: MENA Region Faces 4 Challenges

International Monetary Fund Managing Director Kristalina Georgieva speaks at a news conference during the World Bank/IMF Spring Meetings at the International Monetary Fund (IMF) headquarters in Washington, Thursday, April 13, 2023. (AP)
International Monetary Fund Managing Director Kristalina Georgieva speaks at a news conference during the World Bank/IMF Spring Meetings at the International Monetary Fund (IMF) headquarters in Washington, Thursday, April 13, 2023. (AP)
TT
20

IMF Warns: MENA Region Faces 4 Challenges

International Monetary Fund Managing Director Kristalina Georgieva speaks at a news conference during the World Bank/IMF Spring Meetings at the International Monetary Fund (IMF) headquarters in Washington, Thursday, April 13, 2023. (AP)
International Monetary Fund Managing Director Kristalina Georgieva speaks at a news conference during the World Bank/IMF Spring Meetings at the International Monetary Fund (IMF) headquarters in Washington, Thursday, April 13, 2023. (AP)

Growth of the real GDP in the Middle East and North Africa (MENA) region is projected to slow this year to 3.1 percent from 5.3 percent in the previous year, announced Director of the Middle East and Central Asia Department at the International Monetary Fund (IMF), Jihad Azour.

Azour added that inflation is forecast to remain unchanged at around 15 percent this year before declining modestly in 2024.

In a videoconference attended by Asharq Al-Awsat, Azour explained that the MENA countries face four challenges this year, which are dealing with the effects of inflation, global uncertainty, international financing difficulties, and economic reform developments.

Azour explained that dealing with inflation may require increasing interest rates, which affects economic growth. At the same time, uncertainty and geopolitical tensions pervade all global horizons, and their consequences fall on everyone's shoulders.

Concerning oil-importing countries, the rise in energy prices increases the risks, especially with the increase in the cost of financing and the difficulty in obtaining it. As for the oil-exporting countries, the most critical challenge is growing and diversifying revenues.

Meanwhile, Italian Foreign Minister Antonio Taiani said Thursday that his country wants the IMF to start disbursing a loan to Tunisia without conditions.

During a press conference with his Tunisian counterpart, Tajani vowed to work on Tunisia's behalf in negotiations with the IMF, repeating Italy's proposal that the loan be delivered in two tranches and not be fully dependent on all reforms being in place.

"But not utterly conditional on... the conclusion of the reform process. Start financing, encourage the reforms," he told reporters.

Last week, President Kais Saied rejected IMF "diktats", which asked Tunisia to carry out economic reforms and subsidy cuts as terms for the stalled bailout.

Saeed said he would "not hear diktats" from abroad, warning that the subsidies could lead to unrest.

European leaders feared the collapse of the Tunisian economy could increase the influx of immigrants to European shores.

Tunisia's debts amount to about 80 percent of its gross domestic product, and it reached a preliminary agreement with the Fund in mid-October for a new $1.9 billion loan to help overcome the financial crisis.

However, talks reached a dead end after Tunisia failed to implement a reform program to restructure more than 100 indebted state-owned companies and lift subsidies on some essential goods and services.



HUMAIN to Launch ‘Allam,’ the First Arabic AI Foundation Model from Saudi Arabia

Allam is the first foundational AI model developed from scratch in Saudi Arabia, focusing on the Arabic language and its dialects
Allam is the first foundational AI model developed from scratch in Saudi Arabia, focusing on the Arabic language and its dialects
TT
20

HUMAIN to Launch ‘Allam,’ the First Arabic AI Foundation Model from Saudi Arabia

Allam is the first foundational AI model developed from scratch in Saudi Arabia, focusing on the Arabic language and its dialects
Allam is the first foundational AI model developed from scratch in Saudi Arabia, focusing on the Arabic language and its dialects

In a bold move reflecting Saudi Arabia’s rapidly accelerating digital transformation, tech company HUMAIN is preparing to launch “Allam” - a foundational artificial intelligence model developed and trained entirely within the Kingdom.

Far from being just another addition to the world of large language models, Allam represents a clear statement from the Arab world: it has the capacity to innovate, build, and compete in this critical field on its own terms.

In an exclusive interview with Asharq Al-Awsat, HUMAIN CEO Tareq Amin revealed that the model will debut at the end of August. Allam, he explained, is built from the ground up to focus on the Arabic language in all its forms, from classical Arabic to a wide range of regional dialects, and is equipped with cultural and political safeguards tailored for the region.

“This is not just another large language model,” Amin said. “It’s proof that the Arab world can innovate, train, and deploy AI at a world-class level, according to our own standards.”

A Saudi-Built Innovation

The project was driven by a team of 40 PhD researchers, all based in the Kingdom. Working under tight confidentiality, they built what Amin describes as “the best Arabic model designed to meet our real needs.”

Allam was trained on proprietary datasets that, the company emphasizes, will “never be released on the public internet.” This gives it an unparalleled depth of local knowledge and accuracy in understanding compared with global models.

The model will first be available to the public via HUMAIN Chat, a free Arabic-language application similar to ChatGPT but with key differences. It not only handles formal Arabic with precision but can also converse naturally in dialects such as Saudi, Egyptian, Jordanian, and Lebanese. The system has already been tested in sensitive applications, including Sawtak, a tool for transcribing court session proceedings in Saudi Arabia.

“ChatGPT will never have the datasets we do,” Amin said. “I want the Arab world to start asking: why don’t we build a coalition to create AI models that reflect our culture and values?”

From the outset, Allam was designed to operate within a clear framework of responsible AI. Built-in safeguards at both the input and output stages ensure that its responses align with the cultural, social, and political norms of the region.

“This isn’t about censorship,” Amin stressed. “It’s about relevance and trust. A model is like a child: it needs guidance, education, and refinement to become a responsible adult. That’s our approach with Allam.”

HUMAIN itself is the product of a unique alliance, combining technical expertise from Aramco Digital and Saudi Arabia’s National Center for AI under the Saudi Data and Artificial Intelligence Authority (SDAIA). Amin views the launch not as a finish line, but as the starting point for continuous improvement, driven by feedback from users across the Arab world.

The company’s broader vision is to create a marketplace where developers and businesses can access Allam and deploy ready-made use cases - from business automation to citizen services - without having to start from scratch.

The Size of the Opportunity

Arabic is spoken by more than 350 million people worldwide, yet Amin points out that it remains underrepresented in leading AI models, which are typically trained primarily in English and a small number of other languages. Even when Arabic support is available, coverage of dialects and cultural nuances is limited.

HUMAIN’s focus is therefore squarely on serving government entities that rely almost entirely on Arabic, as well as private-sector industries such as tourism and healthcare.

For Amin, Allam is more than just a linguistic project. “It’s the spark that can shift the Middle East’s position in the global digital economy, from consumer to creator of original platforms and products,” he said. “We don’t yet have a complete AI ecosystem of developers and companies. We need to believe in our abilities, and the time is now.”

World-Class Infrastructure

Alongside Allam, HUMAIN has been investing heavily in infrastructure. The company recently announced a major agreement with Silicon Valley startup Groq, known for its ultra-fast, cost-efficient AI inference technology.

Amin’s relationship with Groq began two years ago when he met CEO Jonathan Ross, the original inventor of Google’s Tensor Processing Units (TPUs), at an event in Saudi Arabia. Impressed by Groq’s ASIC-based architecture optimized for inference, Amin decided to integrate their technology into HUMAIN’s operations.

That bet has paid off. HUMAIN deployed 19,000 Groq Language Processing Units (LPUs) in just six days, enabling inference services at roughly 60% lower cost than anywhere else globally. The system boasts low energy consumption, SRAM-based memory architecture, and a custom design optimized for running large models efficiently.

OpenAI Models Go Live in Saudi Arabia

The HUMAIN –Groq partnership has already delivered a milestone: the immediate availability of OpenAI’s two latest open-source models - gpt-oss-120B and gpt-oss-20B - on the GroqCloud platform, with full local hosting in the Kingdom.

Both models support a 128,000-token context window, provide real-time responses, and include integrated tools such as code execution and web search. Today, HUMAIN’s Groq-powered inference infrastructure in Dammam is serving users in 130 countries, a first for Saudi Arabia, and likely for the Middle East as a whole.

Rethinking the Enterprise Operating System

While Allam is HUMAIN’s flagship model, the company is also gearing up for another major release in October: HUMAIN One, which Amin describes as “a complete reinvention of the enterprise operating system.”

Instead of switching between dozens of separate applications, users interact with a single unified interface - text or voice-based - that can execute complex tasks seamlessly across multiple systems.

In one pilot case, a single AI agent reduced a payroll preparation process from 30 staff-hours involving four employees down to just 30 minutes, with higher accuracy. HUMAIN One’s voice interface will work on Windows, macOS, and HUMAIN’s own AI-enabled PCs, which all company staff currently use.

The HUMAIN AI Computer

This integration will extend to HUMAIN’s own AI computer, designed entirely in Saudi Arabia in partnership with Qualcomm. The device combines CPU, GPU, and Neural Processing Unit (MPU) capabilities for comprehensive AI computing power, tailored for advanced applications.

The HUMAIN AI computer will debut at the Future Investment Initiative (FII) in Riyadh this October, with a global release planned afterward. “It will change the game,” Amin said. “When you see its specs and price compared to the market, you’ll understand our edge computing strategy - delivering fast, efficient local processing without over-reliance on remote data centers.”

AI as an Economic Pillar

From Allam to Groq-powered infrastructure to HUMAIN One, all of HUMAIN’s initiatives align with Saudi Vision 2030. Amin views AI as “the foundation upon which the entire strategy is built”, not only in tourism, healthcare, and industry, but across every sector.

He praised Crown Prince Mohammed bin Salman’s approach as “both visionary and pragmatic,” treating AI “not as an optional tool, but as a necessity for economic growth, citizen empowerment, and sector-wide adoption.”

Investing in Local Talent

For Amin, HUMAIN’s success is first and foremost the result of its people, especially the Kingdom’s deep pool of AI talent.

“Some doubted whether we had the capabilities,” he said. “I told them: come and see for yourself.”

The presence of 40 PhD researchers behind Allam, he argued, is living proof that the Middle East can produce world-class AI models and challenge the assumption that the region must rely on external innovation.