AI News

Follow the latest AI and automation news

23 Aug 202611 Rabiʻ I 1448 AH

AI is becoming AI's biggest customer as agentic token usage jumps 14x on OpenRouter

Since February 6, 2025, AI agents have surpassed human token usage on OpenRouter, with agent usage increasing by 14 times compared to a 2.8 times increase for humans. Data indicates that nearly 70% of agent token consumption comes from inexpensive cached prompts, suggesting that actual costs are rising much more slowly than the raw figures imply. This shift signifies that AI is becoming the largest customer of AI, highlighting the growing importance of agents and indicating potential changes in operational costs and development in the field.

Arab AI Scene

23 Aug 202611 Rabiʻ I 1448 AH

Syira AI launches with Doha office

Syira AI has officially opened its new office in Doha, marking the start of its operations in Qatar. The company offers 'Intelligence-as-a-Service', combining consulting, custom development, and automation. Syira AI targets key sectors including education, healthcare, construction, retail, government, and finance.

Governance

23 Aug 202611 Rabiʻ I 1448 AH

Asia is reaching for the open road that Chinese AI offers

At a recent digital technology and AI forum in Chengdu, ministers from the Asia-Pacific Economic Cooperation issued a statement supporting open-source AI models. This marks a rare consensus in a debate often framed as a zero-sum contest between the US and China, indicating a shift in understanding regarding the future of AI.

💡

Memory shortage reportedly drives Nvidia AI server prices up about 15 percent

How will rising prices affect companies' investments in AI?

Business

Alibaba to issue US$10 billion in new shares for huge AI push amid strong investor demand

23 Aug 202611 Rabiʻ I 1448 AH

Business

23 Aug 202611 Rabiʻ I 1448 AH

Alibaba to issue US$10 billion in new shares for huge AI push amid strong investor demand

Alibaba Group Holding announced it will issue HK$80 billion (US$10.2 billion) in new shares, using all proceeds to invest in AI. This move aims to extend the company's global leadership in AI, with plans to enhance its full-stack AI capabilities and infrastructure. This decision reflects Alibaba's ongoing efforts over recent years to transition from an e-commerce giant to a leader in AI technology. The company is committed to developing comprehensive AI capabilities, showcasing its dedication to innovation in this field. This investment aligns with global trends towards enhancing AI technologies.

Agents

Harvard’s $699 startup bootcamp offers AI avatars of its instructors

22 Aug 202610 Rabiʻ I 1448 AH

Agents

22 Aug 202610 Rabiʻ I 1448 AH

Harvard’s $699 startup bootcamp offers AI avatars of its instructors

The HBS Foundry program employs AI avatars to provide feedback during practice pitches and board meetings. This technology offers a unique interaction, allowing trainees to receive immediate comments on their performance. This approach enhances training effectiveness and helps participants improve their skills more rapidly.

Models

Inherent, founded by DeepMind alumni, says its AI ‘teammate’ just outperformed Anthropic and OpenAI at replicating research

22 Aug 202610 Rabiʻ I 1448 AH

Models

22 Aug 202610 Rabiʻ I 1448 AH

Inherent, founded by DeepMind alumni, says its AI ‘teammate’ just outperformed Anthropic and OpenAI at replicating research

Inherent AI Lab, founded by DeepMind alumni, has announced the launch of Faraday, an AI agent capable of replicating scientific papers. This model could serve as a significant step towards innovation in the field. Inherent aims to push the boundaries of artificial intelligence, with Faraday enabling researchers to accurately replicate studies. This technology may expedite research and development processes, making it particularly relevant in academic and industrial contexts.

Governance

OpenAI says California should strengthen its AI safety bill

22 Aug 202610 Rabiʻ I 1448 AH

Governance

22 Aug 202610 Rabiʻ I 1448 AH

OpenAI says California should strengthen its AI safety bill

OpenAI has called on California to strengthen the AI safety bill SB 53, which it previously opposed. This shift in stance comes as concerns grow regarding the safety of artificial intelligence and its potential impacts. The bill is seen as a crucial step toward regulating AI use and ensuring its safety across various applications. By supporting this bill, OpenAI aims to highlight the need for a legal framework that ensures AI safety, especially as reliance on the technology increases in multiple fields. The bill includes new requirements aimed at assessing the risks associated with AI technology, reflecting the company's change in position.

Governance

Frontier AI labs still won’t say how they’d contain a rogue model

22 Aug 202610 Rabiʻ I 1448 AH

Governance

22 Aug 202610 Rabiʻ I 1448 AH

Frontier AI labs still won’t say how they’d contain a rogue model

A new study reveals that leading AI labs have insufficient documented plans for addressing rogue models. These findings raise concerns about the readiness of these labs to handle unexpected and potentially dangerous behaviors exhibited by their systems, as most have not provided clear strategies for managing these risks. This lack of planning highlights the urgent need for effective standards and procedures to ensure the safe use of AI, impacting the future development of this technology in the market.

Agents

Study explains why AI agents benefit from "skills" and when they fail

22 Aug 202610 Rabiʻ I 1448 AH

Agents

22 Aug 202610 Rabiʻ I 1448 AH

Study explains why AI agents benefit from "skills" and when they fail

A study from Princeton University and UC San Diego found that AI agents improve mainly through structured workflows rather than additional knowledge. This insight highlights the importance of workflow design in enhancing agent performance. As the skill library expands, agents struggle more to find the right set of instructions, which can hinder their efficiency. This growing complexity presents new challenges in the design of intelligent agents. The study emphasizes the need for effective strategies to manage skills in AI agents, which is crucial for improving their future performance. This insight is vital for decision-makers in the AI field.