Cloudflare's 2026 App Innovation Report reveals that application modernization, not AI adoption itself, determines success. Surveying 2,300+ senior leaders globally, the research shows organizations ahead in modernizing applications are nearly 3x more likely to achieve clear AI returns. In APAC, 92% of leaders identify software updates as the most critical factor for improving AI capabilities. While AI budgets and pilots expand, uneven results stem from underlying application infrastructure rather than the technology itself, making modernization the key differentiator between organizations realizing AI value and those struggling to see tangible outcomes.

Citi has integrated AI into daily operations across its 182,000-employee workforce, with over 70% now using approved AI tools. Through its "AI Champions" and "AI Accelerators" programs, the bank has built an internal AI workforce of 4,000 employees from diverse departments including technology, operations, risk, and customer support. Unlike competitors who limit AI to specialists or pilot projects, Citi has spent two years embedding the technology organization-wide, prioritizing widespread participation over centralized control and achieving broader adoption than peers who restrict access to technical teams.

Enterprises considering Chinese open-weight AI models like Moonshot AI's Kimi K3 face uncertainty about long-term viability amid renewed US policy debates. Released as the largest open-weight model in July, K3 sparked Washington discussions on federal procurement rules, export controls, and security advisories that could impact global cloud providers and procurement decisions worldwide, extending far beyond US borders.

OpenAI Presence is a managed enterprise AI agent service launched in July 2024, departing from traditional API-based models. Delivered through limited general availability with Forward Deployed Engineers and systems integrators, it's sold as a project-based solution rather than self-serve product. Each deployment focuses on specific tasks like billing disputes, insurance claims, or IT service requests, with customer-defined rules governing agent permissions, approvals, and human handoffs. Post-launch, the system analyzes production sessions to recommend improvements, offering businesses controlled AI automation for targeted workflows.

ChatGPT now offers a Health feature allowing users 18 and older to seamlessly integrate Apple Health data and medical records directly into conversations. Available across web and iOS for Free, Go, Plus, and Pro tiers, this integration enables users to connect records from US hospital systems, One Medical, and Function Health. Once synced, ChatGPT can access medications, lab results, recent visits, sleep data, and activity logs throughout any conversation, eliminating the need for a separate health section. This design choice came after OpenAI discovered that over 70 percent of health-related discussions occurred naturally during meal planning or symptom queries rather than in dedicated health spaces. The feature transforms how users interact with their health information, making medical data contextually available when needed most, whether discussing nutrition, symptoms, or wellness goals, all while maintaining conversation flow within the familiar ChatGPT interface.

Two dozen leading tech companies including Meta, Microsoft, Nvidia, IBM, and Dell Technologies have signed an open letter urging US policymakers to protect open-weight AI models. The coalition, which also includes Hugging Face, Perplexity, Mistral, and major investors like Andreessen Horowitz, compares the current debate to the 1980s open-source software movement. Open-weight AI models allow users to freely download, inspect, modify, and run trained parameters on their own hardware, contrasting with closed systems from companies like OpenAI and Anthropic that restrict access through commercial APIs. This unprecedented alliance of commercial rivals advocates for keeping AI development open and accessible.

Nvidia's Medical Physics Simulation framework revolutionizes healthcare robotics by treating them as physical AI systems requiring embodied learning through contact, force, and consequences rather than code alone. Unlike language models learning from text, these systems master real-world interactions like catheter navigation and tissue manipulation through detailed computational simulations. This approach addresses the scarcity of real-world medical training scenarios, which are heavily regulated and time-consuming, by manufacturing embodied experiences digitally to accelerate robot learning in healthcare applications.

AMD announces a landmark $5 billion investment in Anthropic as part of a massive infrastructure deal worth tens of billions. Anthropic will deploy up to two gigawatts of AI computing capacity using AMD's next-generation Instinct MI450-series accelerators, with the first gigawatt rollout scheduled for the first half of 2027. This strategic partnership combines AMD's equity commitment with a substantial hardware procurement agreement, where the investment is milestone-based and tied to Anthropic's deployment targets. The deal follows AMD's pattern of linking financial incentives to chip deployments, similar to its October agreement with OpenAI for six gigawatts of capacity. This collaboration represents a significant expansion of AMD's presence in the AI infrastructure market and demonstrates Anthropic's commitment to scaling its AI capabilities with cutting-edge hardware technology.

SenseTime launches the Galaxy Project, partnering with nearly 20 organizations to expand domestic AI chip infrastructure in China. Co-founder Yang Fan outlined a comprehensive strategy connecting chip technology, ecosystem partnerships, and commercial deployment of domestically-produced AI computing power. The initiative includes a space computing agreement with Guoxing Aerospace and research collaborations with five institutions, including Shanghai Artificial Intelligence Laboratory, focusing on scientific computing applications. The project addresses three key trends: rising enterprise token demand, growing industrial AI adoption, and advancing domestic chip commercialization for intelligent computing centers.

Bristol Myers Squibb becomes the first life sciences company to acquire Nvidia's DGX SuperPOD powered by Vera Rubin architecture for AI-driven drug discovery and development. The system features eight DGX Vera Rubin NVL72 rack-scale units combining Vera CPUs and Rubin GPUs. BMS will utilize this advanced infrastructure to train proprietary AI models and execute predictions across research programs involving compounds, proteins, and scientific data, expanding its existing Nvidia computing capabilities to accelerate pharmaceutical innovation.

Discover how generative AI tools designed for efficiency may actually be hurting your productivity. In 2026, knowledge workers are finding that AI chatbots encourage prolonged engagement rather than task completion, pulling focus away from deep, uninterrupted work. Learn why these interfaces are built to keep you in the loop and what that means for how you work.

Discover Google's latest AI models—Gemini 3.6 Flash and 3.5 Flash-Lite—engineered to reduce latency and token costs for enterprise AI agents. Designed for high-volume production environments, these models prioritize throughput and efficiency over raw parameter count, making them ideal for autonomous, multi-step workflows. Learn how Google's strategic model lineup addresses the cost-performance trade-offs critical to scaling AI agent deployments.

Moonshot AI's Kimi K3 is making waves in the AI community as a powerful open-weight model that stands out for its impressive parameter count. Designed to push the boundaries of open-source AI development, Kimi K3 represents Moonshot AI's commitment to advancing accessible and high-performance language models. With its competitive architecture and capabilities, Kimi K3 positions itself as a strong contender in the growing landscape of open-weight AI models, offering researchers and developers a robust solution for a wide range of natural language processing tasks. Discover what makes Kimi K3 a groundbreaking addition to the open-weight model ecosystem.

Discover how the PULSE programme is bringing generative AI to US public health departments. Led by the Coalition for Health AI, OpenAI, Anthropic, and Accenture, this initiative will pilot AI tools across 10 jurisdictions, giving up to 2,000 public health practitioners access to enterprise AI products. Learn how this landmark programme aims to develop real-world implementation guidance for public health agencies considering AI adoption.

Formula E has expanded its multi-year partnership with Google Cloud, integrating Gemini AI models to support net zero targets through smarter logistics, performance analysis, and back-office operations. Unlike traditional sponsorships, this collaboration focuses on real business outcomes, from race management to fan experience. The agreement, elevated to "Principal Partner" status, builds on a technical foundation established in January 2025, leveraging Google Cloud infrastructure to drive efficiency across Formula E's global commercial and operational workflows.

Discover how AI is transforming enterprise network requirements beyond simple speed. In the AI age, connectivity must be always-on, predictable, observable, and resilient. As AI workloads become distributed and latency-sensitive, networks evolve from infrastructure into critical system dependencies — where degradation directly impacts application performance. Learn what it truly means to build an AI-ready network.

Discover how airlines like Air France-KLM are leveraging generative AI to navigate severe weather disruptions and operational challenges. From managing increased customer queries to making rapid safety-compliant decisions, AI is transforming the aviation industry. Learn how cloud-based AI platforms, built in partnership with tech leaders like Accenture and Google Cloud, are helping airlines become more efficient, reactive, and resilient in the face of unprecedented disruptions.

Discover how RavenDB redefines database management by eliminating the traditional trade-offs between performance, flexibility, and security. Founded by Oren Eini, a former database performance consultant who witnessed capable teams struggling with growing system complexity, RavenDB was built to free developers and administrators from forced compromises. Unlike conventional databases that require manual tuning for speed, sacrifice flexibility for structure, or treat security as an afterthought, RavenDB abstracts away complexity to deliver all three core qualities simultaneously. Learn how this innovative approach is transforming the way organizations manage their data infrastructure.

Discover how Google DeepMind and Isomorphic Labs are tackling AI biosecurity risks through their bioresilience program. With over 15 partnerships across government bodies, biosecurity organizations, and research groups, the initiative addresses the dual-use challenge of frontier AI models like Gemini in biology. While these powerful tools accelerate vaccine research and outbreak response, they also pose potential misuse risks. Learn how DeepMind and Isomorphic Labs are working to enable groundbreaking scientific advances while preventing malicious actors from exploiting AI-driven biological knowledge.

Bunkerhill Health has secured $55 million in Series B funding to expand Carebricks, its agentic AI platform designed for real-world clinical deployment. Backed by Sequoia Capital, Khosla Ventures, Felicis, Optum Ventures, and Y Combinator, the company is addressing a critical gap in healthcare AI — bridging the divide between sandbox models and live clinical environments. Unlike many AI pilots that fail to reach actual patient care, Carebricks is built to operate against live clinical data at institutional scale. With US healthcare spending reaching $5.3 trillion in 2024 and persistent staffing challenges, Bunkerhill Health positions Carebricks as a practical, deployable solution for health systems seeking proven AI performance beyond the research setting.

Anthropic has partnered with the UK's Department for Science, Innovation, and Technology (DSIT) to develop advanced government AI assistant capabilities, building on their February 2025 Memorandum of Understanding. Unlike traditional chatbots, the initiative focuses on deploying agentic AI systems designed to actively guide citizens through complex government processes rather than simply providing static information. This move addresses a critical gap in digital service delivery, where data-rich government portals often require specialist knowledge that many users lack. By operationalising AI beyond the proof-of-concept stage, the project aims to modernise how citizens interact with state services, making them more accessible and easier to navigate for everyone.

Neko Health has secured $700 million in Series C funding to expand its AI-powered preventive body scan services to the United States, beginning with a New York clinic. Led by Lightspeed Venture Partners and O.G. Venture Partners, the round brings the company's total disclosed funding since 2023 to over $1 billion, following a $65 million Series A in 2023 and $260 million raised in January 2025. The service integrates medical imaging, blood tests, proprietary sensors, and clinician review to deliver comprehensive health screenings. Notable investors include Mark Zuckerberg, Priscilla Chan, Maria Sharapova, will.i.am, and Thierry Henry. With strong institutional and high-profile individual backing, Neko Health is poised to reshape preventive healthcare through advanced AI diagnostics.

Discover the top enterprise AI security tools designed to protect AI-driven workflows, proprietary data, and automated systems. From AI discovery and governance to runtime hardening, red teaming, and SOC alert management, this guide breaks down the key functional categories of AI security solutions helping organizations safely scale AI adoption. Whether you're securing LLM applications, agent pipelines, or SaaS integrations, find the right tools to manage risk across your entire AI stack.

Discover Nokia's AI-RAN platform, launched on July 15 and claimed to be the industry's first. Built on Nokia's anyRAN software and NVIDIA's Aerial system, this groundbreaking platform enables operators to maximize capacity from existing spectrum. Described as one of the most significant shifts in radio architecture in decades, Nokia's AI-RAN promises to revolutionize how telecom operators manage and optimize their network resources.

Standard Chartered is redefining how global banks approach AI development by placing data privacy at the core of every stage — from design to deployment. Operating across multiple jurisdictions with varying privacy regulations, the bank faces complex challenges around data usage, storage permissions, and accountability. To navigate this, privacy teams have taken a central role in shaping AI system design, approval, and monitoring processes. According to David Hardoon, Global Head of AI Enablement at Standard Chartered, data privacy functions have become the foundation of most AI regulations, directly influencing what data can be used, how transparent AI systems must be, and who holds responsibility once they go live. This privacy-first approach reflects a broader shift in the financial industry, where regulatory compliance and responsible AI governance are no longer afterthoughts but essential starting points for any AI initiative.

Bluesight's Prism is an AI layer built on AWS that connects hospital pharmacy and compliance data across its product suite. Prism Assistant for ControlCheck is now generally available across 20 health systems, automating the review of 340B GPO drug purchase exceptions. By integrating FDA shortage notices, ASHP records, inventory data, and ML forecasts, Prism reduces the over 4,000 annual staff hours hospitals spend on manual compliance reviews, with a multi-product 340B GPO compliance agent planned for release in 2026.

Artificial intelligence is reshaping the US workplace, but adoption remains uneven across industries and roles. According to a Gallup Workforce survey of over 23,000 US adults conducted in August 2025, AI use is growing yet far from universal. Knowledge workers in technology, finance, and professional services lead adoption, with over 75% of IT professionals using AI regularly. Discover key insights into how employees engage with AI, who benefits most, and where gaps remain.

Cloudflare is reshaping how AI bots access the web with a major policy update. Starting September 15, AI agent crawlers and training bots will be blocked by default on ad-supported pages for newly onboarded domains. The platform has replaced its single AI-bot toggle with three distinct categories: Search, Agent, and Training — giving website owners granular control over which automated systems can access their content. Agent crawlers, which fetch pages in real time for users via tools like ChatGPT, and Training crawlers that harvest data for model development, will face default restrictions, while Search indexing bots remain permitted. These controls have been available to all Cloudflare customers, including free-tier users, since July 1. The update signals a significant shift in how publishers can protect their content from AI systems while maintaining search visibility.

Explore how the White House paper "Artificial Intelligence and the Great Divergence" draws compelling parallels between AI and the Industrial Revolution. Discover how AI now drives US economic strategy, boosting GDP by 1.3% in early 2025 through major infrastructure investments like data centres. Learn about projections showing AI could deliver productivity growth of up to 20%, positioning artificial intelligence as the defining force shaping global economies for decades to come.

Discover how forward-thinking organisations are preparing for the rise of AI agents while navigating critical governance challenges. With IDC predicting that by 2030, up to 20% of Global 1000 companies could face lawsuits and fines due to inadequate AI controls, the need for robust guardrails has never been greater. Enter the Masumi Network — a groundbreaking initiative born from a collaboration between NMKR and Serviceplan Group. Launched in late 2024, this framework-agnostic infrastructure combines agentic AI with decentralisation to help developers build safer, more accountable AI ecosystems. Learn how blockchain technology is shaping the future of AI governance.

Discover why scaling enterprise AI goes beyond model selection. Franny Hsiao, EMEA Leader of AI Architects at Salesforce, reveals the architectural oversights stalling AI pilots before production and how organisations can build reliable, real-world AI systems through smarter data engineering and governance strategies. Ahead of AI & Big Data Global 2026 in London.

Discover how agentic AI is transforming business intelligence and analytics. ThoughtSpot's field chief data and AI officer Jane Smith explains how agentic systems are shifting organizations from passive reporting to active, real-time decision-making. Unlike traditional BI tools that wait for user queries, agentic AI proactively monitors data from multiple sources around the clock, diagnoses changes, and automatically triggers next steps. Learn how data and analytics leaders can navigate this unprecedented wave of change and reimagine their BI strategies from the ground up with the help of innovative platforms like ThoughtSpot.

HMRC has selected SAP to modernise its core tax administration systems, placing AI at the heart of the UK's revenue management strategy. The overhaul targets the Enterprise Tax Management Platform (ETMP), which manages over £800 billion in annual tax revenue across 45+ tax regimes. By migrating to a managed cloud environment through RISE with SAP, HMRC aims to replace legacy infrastructure with a unified architecture that natively supports machine learning and automated decision-making. This shift enables the consolidated data sets required for effective AI deployment, moving beyond simply layering tools onto outdated systems. The transformation will impact tens of thousands of staff and signals a significant turning point in how public sector organisations approach intelligent automation and digital modernisation at scale.

Nvidia CEO Jensen Huang has set a bold benchmark for AI adoption: engineers should spend at least half their salary — around $250,000 — on AI tokens annually, or risk raising red flags. Huang revealed Nvidia is targeting a $2 billion yearly token bill across its engineering workforce. This reflects a broader industry shift where AI spending is replacing human labor costs. Major hyperscalers are projected to spend a combined $700 billion in 2026 capital expenditure — nearly double last year — while AI remains the top reason cited for U.S. job cuts for a record fourth consecutive month. Companies like Meta have explicitly linked layoffs to offsetting AI investment costs. The message is clear: in the modern tech economy, token budgets are becoming as critical as headcount, and the engineers who leverage AI most aggressively are the ones companies want to keep.

Discover how NHS hospitals are adopting PinPoint's AI-powered blood test to revolutionize womb cancer screening. Developed by Leeds-based PinPoint Data Science, this innovative £30 test uses machine learning to analyze around 30 blood markers, classifying patients as low, elevated, or high risk. With 90,000 postmenopausal women referred annually in England for potential womb cancer checks, this non-invasive tool aims to streamline referral pathways, reduce unnecessary procedures, and prioritize faster assessment for high-risk patients. Learn how this breakthrough multi-cancer test is transforming early detection and improving outcomes for women across the UK.

Discover how AWS GraphRAG is transforming pharmaceutical research by reducing drug development cycles by 87%. By integrating isolated proprietary databases into a unified knowledge graph powered by Amazon Neptune Analytics and Bedrock, researchers can now uncover hidden data correlations through simple natural language queries. What once took over six months per iteration with only a 5% success rate is now streamlined through intelligent graph-based retrieval and NLP technology — accelerating breakthroughs and preserving critical project knowledge across teams.