AI News

Follow the latest AI and automation news

10 Sept 202629 Rabiʻ I 1448 AH

Google Cloud and Accenture expand their long-term partnership

Accenture and Google Cloud have announced an expansion of their partnership with a new global initiative aimed at assisting enterprise clients in deploying AI tools. The newly formed Accenture Gemini Enterprise Business Group will integrate Accenture's certified talent with Google Cloud engineering resources to help clients transition AI projects from experimental stages to commercial production.

Business

10 Sept 202629 Rabiʻ I 1448 AH

Kimi maker Moonshot AI to explore dual Hong Kong and Shanghai IPOs to boost capital: sources

Moonshot AI, the Chinese start-up behind the Kimi K3 model, is considering dual listings in Hong Kong and Shanghai to secure more capital. Following a meeting with its financial backers in July, the Beijing-based company has begun negotiating detailed arrangements for an initial public offering in Hong Kong, while hinting at a potential listing on a mainland exchange afterwards. This dual listing is a strategic move to enhance its market presence and secure necessary funding for expansion, reflecting a growing trend among Chinese firms seeking global financing options.

Business

10 Sept 202629 Rabiʻ I 1448 AH

Warnings About AI Risks to Global Financial Stability

Pablo Hernandez de Cos, head of the Bank for International Settlements, warned that the rapid rise of artificial intelligence creates new risks to financial stability, noting that spending on related infrastructure has become significant enough to impact global economic conditions. The Bank for International Settlements estimates that the top five tech companies will spend over one trillion dollars on AI in 2025 and 2026, while sector forecasts suggest global AI investments could rise from around 500 billion dollars currently to as much as four trillion dollars by 2030.

💡

DeepSeek Launches New DeepSeek-V4.1-Flash AI Model

How will the launch of DeepSeek-V4.1-Flash affect the AI market?

Business

Expanding AI access and cyber defense for federal, state, local, and tribal governments

10 Sept 202629 Rabiʻ I 1448 AH

Business

10 Sept 202629 Rabiʻ I 1448 AH

Expanding AI access and cyber defense for federal, state, local, and tribal governments

OpenAI has announced a government support initiative for federal, state, local, and tribal entities, offering $0 license fees for eligible organizations. Additionally, the company will provide a 50% discount on usage, facilitating access to its advanced technologies for government agencies. This initiative also includes expanded cyber defense support, demonstrating OpenAI's commitment to enhancing cybersecurity for government entities.

Business

Introducing ChatGPT for Financial Services

10 Sept 202629 Rabiʻ I 1448 AH

Business

10 Sept 202629 Rabiʻ I 1448 AH

Introducing ChatGPT for Financial Services

OpenAI has announced the launch of ChatGPT for Financial Services, which combines integrated financial data with the GPT-6 Astra model. This new system aims to enhance research, modeling, and client-ready materials. ChatGPT for Financial Services features access to real-time financial data, improving the accuracy of financial analyses and models. This innovation marks a significant step in integrating AI into the financial sector, enabling professionals to access precise and timely information.

Arab AI Scene

KAUST Celebrates 290 Graduates of Specializations Program

10 Sept 202629 Rabiʻ I 1448 AH

Arab AI Scene

10 Sept 202629 Rabiʻ I 1448 AH

KAUST Celebrates 290 Graduates of Specializations Program

King Abdullah University of Science and Technology (KAUST) celebrated the graduation of 290 students from its Specializations Program, aimed at developing Saudi talent in scientific and technological fields of national priority. The program covers five core specializations: artificial intelligence, bioinformatics, cybersecurity, renewable energy, and water sustainability. It attracted significant interest across the Kingdom, with 23,995 applicants, of whom 5,065 progressed to the placement test, 2,473 enrolled in the Introduction Course, and 1,176 qualified for the Advanced Course.

Models

Alibaba Opens Qwen3.8-2.4T-A95B Weights as First Qwen-Max-Class MoE

10 Sept 202629 Rabiʻ I 1448 AH

Models

10 Sept 202629 Rabiʻ I 1448 AH

Alibaba Opens Qwen3.8-2.4T-A95B Weights as First Qwen-Max-Class MoE

Alibaba's Qwen team has announced the release of open weights for the Qwen3.8-2.4T-A95B model, marking the first Qwen-Max-class model available for download instead of API access. The model features approximately 2.4 trillion parameters, activating about 95 billion per token across 512 experts, with 10 routed experts and one shared expert used at each step. This release positions a frontier-scale text MoE, making it suitable for self-hosted AI applications.

Models

Ant Group Opens Ling-3.0-flash-VL Multimodal Model With Vision Feedback Loop

10 Sept 202629 Rabiʻ I 1448 AH

Models

10 Sept 202629 Rabiʻ I 1448 AH

Ant Group Opens Ling-3.0-flash-VL Multimodal Model With Vision Feedback Loop

Ant Group has launched the Ling-3.0-flash-VL model, the first multimodal model in its Ling series, through the inclusionAI organization on Hugging Face and ModelScope. Built on the Ling-3.0-flash architecture, it features 124 billion parameters, activating about 5.5 billion per token, and accepts inputs from images, text, and video. The model's design includes a vision feedback loop that treats seeing as part of an ongoing work cycle, making it suitable for GUI automation and medical report generation. In image-to-webpage tests, it demonstrated the ability to recover layout and component relationships, surpassing GPT-5.4's performance.

Agents

AgiBot Unveils GE-Act 2.0 Native World-Action Model With 100x Data Scaling

10 Sept 202629 Rabiʻ I 1448 AH

Agents

10 Sept 202629 Rabiʻ I 1448 AH

AgiBot Unveils GE-Act 2.0 Native World-Action Model With 100x Data Scaling

AgiBot has launched GE-Act 2.0, a world-action model trained entirely from scratch on embodied manipulation data. This release represents the first systematic test to determine if a world-action model can unlock new robot skills as training hours increase from 300 to 30,000. GE-Act 2.0 trains vision representation, future prediction, and action prediction together on robot data, enhancing the model's effectiveness. During testing, success rates improved with data scale, rising from 17.1% to 44.1% on the G1-OP platform.