AI News

Follow the latest AI and automation news

3 Sept 202622 Rabiʻ I 1448 AH

Safety overview: GPT-6 Astra

The company announced the launch of GPT-6 Astra, its most capable model to date. This model achieves a Critical level of cybersecurity capability under the established Preparedness Framework, positioning it as a leader in cybersecurity. This launch represents a strategic step towards enhancing security capabilities in intelligent models, allowing organizations to use it with greater confidence in sensitive environments.

Arab AI Scene

3 Sept 202622 Rabiʻ I 1448 AH

HUMAIN and KORA Partner to Build Operating

On September 3, 2026, HUMAIN, a PIF-owned company, and KORA Systems announced a strategic partnership at LEAP 2026 to develop HUMAIN OS, an AI-native Linux operating system. This collaboration combines HUMAIN's extensive AI capabilities with KORA's expertise in open-source operating systems, aiming to create an enterprise-grade platform.

Models

2 Sept 202621 Rabiʻ I 1448 AH

OpenAI’s new reasoning technique alarms AI safety experts

OpenAI has announced the launch of its new Astra model, which employs a technique called recurrent depth. This technique allows the model to operate outside the sequential thinking that characterizes most reasoning models. Recurrent depth is considered an advanced step in AI model development, enhancing the model's ability to process information more effectively. This innovation could change how models interact with complex data.

💡

Google says its new Gemini 3.8 Flash model ‘works harder’ but might cost more

What impact will Gemini 3.8 have on token usage for developers?

Arab AI Scene

Can AMD's Saudi AI Buildout Challenge NVIDIA and Broadcom?

2 Sept 202621 Rabiʻ I 1448 AH

Arab AI Scene

2 Sept 202621 Rabiʻ I 1448 AH

Can AMD's Saudi AI Buildout Challenge NVIDIA and Broadcom?

AMD is aiming to strengthen its presence in the Saudi AI market through new investments. This move comes at a time when the Saudi market is experiencing significant growth in AI technologies, providing substantial opportunities to compete with major companies like NVIDIA and Broadcom. These investments are part of AMD's strategy to expand its operations in the Middle East, focusing on meeting local market needs and delivering innovative solutions. These developments are expected to positively impact local companies, enhancing their competitiveness in the global market.

Models

These Russian Mathematicians Taught AI Models How to Talk to Each Other Without Using Words

2 Sept 202621 Rabiʻ I 1448 AH

Models

2 Sept 202621 Rabiʻ I 1448 AH

These Russian Mathematicians Taught AI Models How to Talk to Each Other Without Using Words

Mostik, a startup, has announced a novel approach that combines the capabilities of AI models in unconventional ways. This initiative aims to enhance performance and efficiency in data processing. Such a move is significant amid increasing competition in the AI sector, as companies strive to deliver innovative solutions. Mostik is working on integrating multiple AI models, allowing for more accurate and effective results. This innovation aligns with current market trends, where the demand for integrated AI solutions is rising. The startup is also leveraging modern technologies to boost its capabilities.

Agents

Anthropic Returns to Cooperation with Trump Administration on AI

2 Sept 202621 Rabiʻ I 1448 AH

Agents

2 Sept 202621 Rabiʻ I 1448 AH

Anthropic Returns to Cooperation with Trump Administration on AI

U.S. Commerce Secretary Howard Lutnick stated that AI company Anthropic has 'returned to the right side' with the Trump administration, indicating a potential easing of tensions regarding its military technology use. Defense Secretary Pete Hegseth had previously barred Anthropic from securing certain military contracts after the company refused to allow its AI models, Claude, to be used for internal surveillance or autonomous weapons development. Anthropic responded by filing a lawsuit in California to challenge these restrictions.

Governance

US government sides with OpenAI on issue of training LLMs on copyrighted material

2 Sept 202621 Rabiʻ I 1448 AH

Governance

2 Sept 202621 Rabiʻ I 1448 AH

US government sides with OpenAI on issue of training LLMs on copyrighted material

A recent brief indicates that the United States is focused on enhancing its artificial intelligence industry, aiming to establish global standards for AI use. This initiative is part of a broader strategy to maintain leadership in the field, with the U.S. striving to be at the forefront of AI technology development. This underscores the significance of AI in the global economy and suggests that the U.S. aims to set new standards that will influence how AI is utilized worldwide.

Models

Gemini 3.8 Flash is Google's third budget model in six weeks while frontier models remain MIA

2 Sept 202621 Rabiʻ I 1448 AH

Models

2 Sept 202621 Rabiʻ I 1448 AH

Gemini 3.8 Flash is Google's third budget model in six weeks while frontier models remain MIA

Google has launched the Gemini 3.8 Flash model, the third release in six weeks, which matches the Claude Opus 5 on certain coding benchmarks. While Gemini 3.8 Flash is priced lower, it consumes about 30% more output tokens per task, making it more expensive in practice than its predecessor despite having the same token rates. This release comes at a time when frontier models are still missing, reflecting Google's strategy to offer cost-effective models while maintaining performance levels.

Agents

UAE Launches AI Advisor System to Enhance Government Decision-Making

2 Sept 202621 Rabiʻ I 1448 AH

Agents

2 Sept 202621 Rabiʻ I 1448 AH

UAE Launches AI Advisor System to Enhance Government Decision-Making

The UAE Cabinet has approved the 'AI Advisor System for the Cabinet', chaired by Sheikh Mohammed bin Rashid Al Maktoum, to support government decision-making. The system aims to expedite policy and project studies through 32 AI assistants specializing in analyzing legislation and providing recommendations. The UAE seeks to convert 50% of government services to self-executing AI models within two years, redesigning policies and governmental processes. This new system enhances government work efficiency and supports decision-making by providing accurate analyses and continuous monitoring of decisions.