Latest AI News

Amazon Web Services (AWS) and Humain, Saudi Arabia's newly created AI company chaired by Crown Prince Mohammed bin Salman, have announced a strategic partnership to invest over $5 billion in building an 'AI Zone' in Saudi Arabia. The groundbreaking initiative will include dedicated AWS AI infrastructure, advanced servers with cutting-edge semiconductors, and specialized training programs aimed at positioning the Kingdom as a global AI leader. This partnership represents one of the largest international AI infrastructure investments and is expected to significantly accelerate AI development across the Middle East region.

OpenAI's ChatGPT has doubled its weekly active users from 400 million in February to 800 million in May 2025, according to recent data. This explosive growth coincides with the launch of advanced models like GPT-4o, which introduced powerful multimodal capabilities. Despite this success, OpenAI faces intensifying competition from Chinese AI companies like DeepSeek, whose models match or outperform OpenAI's offerings at significantly lower costs.

OpenAI announced a modified restructuring plan that will keep its nonprofit entity in control while converting its for-profit arm into a public benefit corporation, a move that has received SoftBank's endorsement. The Japanese investment giant confirmed its $30 billion commitment remains intact despite the change in restructuring approach. However, Microsoft, another major investor, has yet to approve the plan and is currently renegotiating terms of its multibillion-dollar partnership with OpenAI.

OpenAI plans to reduce Microsoft's revenue share from 20% to approximately 10% by the end of this decade, according to financial documents cited by The Information. This development comes as OpenAI revises its restructuring plans, opting to maintain nonprofit control over its for-profit arm, which will become a public benefit corporation. Microsoft, having invested over $13 billion in OpenAI, is currently negotiating terms to protect its substantial investment while potentially extending its access to OpenAI's technology beyond the current 2030 agreement.

Google has launched the AI Futures Fund, a new initiative designed to invest in and collaborate with startups building with Google DeepMind's advanced AI models. The fund provides early access to cutting-edge AI models like Gemini, Imagen, and Veo, along with technical expertise, cloud credits, and potential direct investment. Unlike traditional accelerator programs, the fund evaluates opportunities on a rolling basis, supporting companies from seed to late stage across various sectors.

OpenAI has launched a new family of models called GPT-4.1, including standard, mini, and nano versions, all optimized for coding and instruction following with a massive 1-million-token context window. These API-exclusive models outperform OpenAI's previous offerings on coding benchmarks, with the flagship GPT-4.1 showing a 21% improvement over GPT-4o. This release comes as competition intensifies with Google's Gemini 2.5 Pro and Anthropic's Claude 3.7 Sonnet also vying for dominance in the AI coding space.

Anthropic's Claude 3.7 Sonnet has set new benchmarks in AI coding performance just two weeks after its release, achieving industry-leading scores on software engineering tests. The model, which combines quick responses with extended reasoning capabilities, is being rapidly adopted by enterprises seeking to accelerate development cycles. Alongside the model, Anthropic launched Claude Code, a command-line AI agent that helps developers build applications faster by handling complex coding tasks.

Las Vegas-based TensorWave has raised $100 million in Series A funding led by Magnetar and AMD Ventures to scale its AMD-powered AI cloud infrastructure. The investment, which brings the company's total capital raised to $146.7 million, will support the expansion of its recently deployed 8,192 AMD Instinct MI325X GPU cluster. As demand for AI compute continues to outpace supply, TensorWave aims to democratize access to high-performance AI infrastructure with its AMD-focused approach.

The Vector Institute has awarded $17,500 scholarships to 120 top AI graduate students across Ontario universities for the 2025-26 academic year. This eighth round of the Vector Scholarship in Artificial Intelligence (VSAI) program represents a $2.1 million investment in Ontario's growing AI talent pipeline. Since launching in 2018, the program has awarded 802 scholarships and expanded to include 28 Vector-recognized master's programs across the province.

Researchers at the University at Buffalo have developed an artificial intelligence system that analyzes children's handwriting to detect early signs of dyslexia and dysgraphia. The technology, presented in the journal SN Computer Science, aims to streamline screening for these learning disabilities by identifying specific patterns in handwriting. This innovation could help address the nationwide shortage of speech-language pathologists and occupational therapists, making early detection more accessible, especially in underserved areas.

Researchers at the University of Tokyo have developed a groundbreaking digital laboratory (dLab) system that fully automates the synthesis and evaluation of thin-film materials. The system leverages machine learning and robotics to autonomously create samples and perform comprehensive measurements without human intervention. By standardizing data formats and connecting modular instruments, dLab accelerates materials development while freeing researchers to focus on creative aspects of scientific discovery.

Stanford University's 2025 AI Index reveals a complex picture of artificial intelligence development, with training costs for frontier models like Google's Gemini Ultra reaching an estimated $192 million. Despite these rising training expenses, the report highlights positive trends: hardware costs have fallen 30% annually, energy efficiency has improved 40% yearly, and inference costs have plummeted 280-fold in just 18 months. However, the environmental impact remains concerning, with models like Meta's Llama 3.1 generating nearly 9,000 tons of carbon emissions during training.

Data analytics giant Databricks has acquired serverless Postgres database provider Neon for $1 billion, marking its third major acquisition in two years. The strategic purchase aims to enhance Databricks' AI agent capabilities, with Neon's technology specifically designed for automated workflows. Internal data revealed that over 80% of databases provisioned on Neon were created by AI agents rather than humans, highlighting the explosive growth of agentic workloads.

Google is set to unveil a sophisticated AI agent designed to assist software engineers throughout the entire development lifecycle at its annual I/O conference on May 20. Known internally as a "software development lifecycle agent," the tool aims to help developers with everything from responding to tasks to documenting code. This move comes as Google faces mounting pressure from investors to demonstrate returns on its substantial AI investments amid intensifying competition in the industry.

Cisco Systems announced on May 14, 2025, that it has raised its annual revenue forecast to between $56.5 billion and $56.7 billion, citing strong demand for AI infrastructure from cloud customers. The company also revealed that its finance chief Scott Herren will retire in July, to be replaced by Mark Patterson, Cisco's current chief strategy officer. Shares of the San Jose-based networking giant rose 2% in extended trading as investors responded positively to the company's AI-driven growth trajectory.

Nvidia-backed AI infrastructure provider CoreWeave has secured an additional $4 billion agreement with OpenAI, as revealed in a regulatory filing on May 15, 2025. This new commitment extends through April 2029 and builds upon their existing $11.9 billion five-year contract signed in March. The announcement reversed CoreWeave's initial stock decline following its first post-IPO earnings report, which had revealed ambitious capital expenditure plans.

MIT researchers have discovered that vision-language models (VLMs) cannot understand negation words like 'no' and 'not', performing no better than random guessing in tests. This fundamental flaw could lead to serious diagnostic errors in healthcare settings, where distinguishing between present and absent conditions is critical. The research team, led by Kumail Alhamoud and Marzyeh Ghassemi, has developed a benchmark called NegBench to evaluate and improve these models.

Scientists at the University of Bristol have unveiled a revolutionary soft robot that mimics an octopus's nervous system, capable of sensing its environment and making decisions without a central computer. The innovative design uses fluid flows of air or water to coordinate suction and movement, similar to how octopuses use hundreds of suckers across multiple arms. This breakthrough demonstrates how suction flow can be used not just for adhesion but also for environmental sensing and autonomous control.

AI startup Cohere has doubled its annualized revenue to $100 million since early 2025, driven by growing demand for secure, customized AI solutions among enterprise clients in regulated sectors. The company's strategic pivot toward private deployments has proven successful, with 85% of its business now coming from these tailored solutions that offer impressive 80% profit margins. This shift reflects a broader industry trend away from large, generalized AI models toward specialized tools designed for specific business needs.

Microsoft has laid off approximately 2,000 employees in Washington state, with software engineers bearing the brunt of these cuts as they represent over 40% of those affected. These layoffs are part of a larger reduction of about 6,000 employees across Microsoft globally, coming shortly after CEO Satya Nadella revealed that AI now generates up to 30% of the company's code. The tech giant declined to comment on whether the rise of AI-assisted coding directly motivated the workforce reduction.