Always-On AI Arrives: Healthcare, Coding, and Robotics Join the Agent Era

What happened
CVS Health announced Health100, an AI‑enabled healthcare engagement platform built with Google Cloud. The system will bring together data from pharmacies, insurers, benefit managers and providers into a unified consumer‑facing app. It will rely on agentic AI — designed to operate with minimal human intervention — to provide “an always‑on personal health care partner” through voice and visual interactions. Health100 will run on Google Cloud’s enterprise infrastructure and use models such as Gemini along with tools like BigQuery and the Cloud Healthcare API. CVS executives said the platform will offer faster care options, help reduce out‑of‑pocket costs and securely manage sensitive health data.

Why it matters
This is one of the most ambitious examples yet of agentic AI moving into healthcare. By unifying disparate data sources and integrating large language models with digital health workflows, CVS aims to deliver personalised, proactive support rather than reactive responses. If successful, it could redefine how consumers interact with healthcare services and demonstrate that agentic AI can handle complex, regulated environments.

What’s next
Health100 will debut later in 2026, with more details promised at Google’s The Check Up event. Expect competing healthcare providers to announce similar agent‑driven platforms, and watch for regulators to scrutinise how such systems handle patient privacy and clinical decision‑making.

What happened
A federal judge in Los Angeles denied Elon Musk’s xAI request to halt enforcement of California’s AI data‑disclosure law, which requires generative AI companies to publish summaries of the datasets used to train their models. Judge Jesús Bernal ruled that xAI had not shown the law likely violates the U.S. Constitution’s free‑speech protections or is otherwise unconstitutional. The law, enacted in September 2024 and effective since January 1 2025, is part of California’s broader push to regulate AI companies and increase transparency. xAI argued that the disclosure requirements would reveal trade secrets, but the court rejected its request for a preliminary injunction.

Why it matters
The ruling underscores the growing tension between AI developers’ desire to protect proprietary data and regulators’ demands for transparency. California’s law could become a template for other jurisdictions seeking to ensure accountability in AI training. If upheld, generative‑AI companies may need to balance intellectual‑property concerns with public and governmental expectations of openness.

What’s next
xAI can continue its lawsuit, but for now the disclosure law remains in force. AI companies should prepare for similar requirements elsewhere and may need to develop strategies for summarising training data without exposing sensitive information. Broader debates over trade‑secret protection and open science in AI are likely to intensify.

What happened
Chile’s state‑owned copper giant Codelco signed a memorandum of understanding with Microsoft to explore joint initiatives in artificial intelligence, advanced analytics, automation and digital security. The 18‑month agreement will involve joint governance and early testing of new solutions, focusing on using intensive data and AI for decision‑making, autonomous operations and automation of critical processes. Codelco CEO Rubén Alvarado said partnering with Microsoft consolidates the miner’s leadership in the future of mining, while Microsoft Latin America president Tito Arciniega emphasised that the collaboration aims to make operations safer, more efficient and sustainable.

Why it matters
AI is moving from proof‑of‑concept to core infrastructure in heavy industry. By joining forces with a global technology provider, Codelco signals that AI‑driven autonomous operations and analytics are critical to the future of mining. Such partnerships could accelerate the deployment of robotic systems, predictive maintenance and sensor‑driven monitoring in extraction industries, improving productivity while aiming to reduce environmental impact.

What’s next
The partnership will run for 18 months, during which Codelco and Microsoft will pilot new AI solutions and share best practices. If successful, the collaboration could lead to wider adoption of AI‑powered automation across the mining sector and inspire similar alliances in other resource‑intensive industries.

What happened
Luma unveiled “Luma Agents,” a new suite of autonomous agents capable of handling end-to-end creative work across text, image, video, and audio. These agents are powered by Luma’s new “Unified Intelligence” models, which can plan, generate, and refine multimodal content, coordinate with other leading AI models, and maintain persistent context across projects. Luma claims its agents recently transformed a $15M, year-long ad campaign into localized versions for multiple countries in just 40 hours for under $20,000, passing all internal quality checks.

Why it matters
This marks a leap in agentic AI orchestration, enabling creative teams and enterprises to automate complex, multi-format campaigns at unprecedented speed and scale. The persistent context and self-critique features set a new bar for agent autonomy and reliability.

What’s next
Expect rapid adoption in marketing, design, and media, with Luma’s orchestration model likely to influence future agent frameworks.

What happened
Amazon launched “Amazon Connect Health,” a suite of healthcare-focused AI agents that automate appointment scheduling, patient verification, medical documentation, and even generate diagnosis and billing codes. The platform is designed to assist both patients and providers, aiming to streamline administrative burdens in clinics and hospitals

Why it matters
This is a major push for agentic AI in healthcare, targeting one of the most paperwork-heavy sectors. By automating routine tasks, Amazon is positioning itself as a key player in healthcare AI infrastructure.

What’s next
Watch for integration with major health systems and potential regulatory scrutiny as AI takes on more clinical documentation roles.

What happened
OpenAI announced GPT-5.4, its new flagship model designed specifically for agentic use cases. GPT-5.4 can control a user’s PC—clicking, typing, and executing multi-step workflows—making it ideal for autonomous agents and complex task automation. The model also brings improved reasoning and workflow management to ChatGPT

Why it matters
This release cements OpenAI’s lead in agentic LLMs, enabling more sophisticated digital coworkers and workflow automation. The ability to directly control devices blurs the line between digital and physical task execution.

What’s next
Expect a surge in agent-powered productivity tools and new debates over security and user consent as LLMs gain more control over devices.

What happened
Cursor, the fast-growing AI coding company, is facing a new reality: with agentic LLMs like GPT-5.4 able to control entire coding workflows, developers may soon no longer need traditional code editors at all. Cursor is rapidly evolving its platform to stay ahead.

Why it matters
The rise of agentic coding assistants could disrupt the entire developer toolchain, shifting value from editors to orchestration and workflow automation.

What’s next
Expect more “agent-first” coding platforms and a shakeup in how software is built and maintained.

What happened
Scientists demonstrated reinforcement learning on photonic chips that compute using optical spikes instead of electronic signals, enabling robotic learning with minimal electricity.

Why it matters
This breakthrough could lead to ultra-efficient, battery-free AI systems for robotics and edge devices, reducing energy costs and expanding deployment options.

What’s next
Look for early prototypes in mobile robots and IoT, and further research into scalable photonic AI hardware.

What happened
China’s latest five-year policy blueprint, released March 5th, sets aggressive targets for AI adoption across all sectors and pledges to accelerate technological self-reliance, especially in semiconductors and emerging tech.

Why it matters
This signals a national commitment to AI leadership and supply chain independence, with major implications for global competition and tech policy.

What’s next
Expect increased investment in domestic AI startups, chipmakers, and research, as well as new export controls and international tensions.

What happened
Apptronik, a robotics startup, has surged in value from $15 million to $5 billion and now counts Google DeepMind as a partner, reflecting the massive capital and strategic interest flowing into humanoid robotics.

Why it matters
The robotics sector is attracting record investment and top-tier AI partnerships, accelerating the path from prototype to commercial deployment.

What’s next
Watch for new product launches, expanded pilots, and further consolidation as the humanoid robotics market heats up.

Agentic AI is moving from experimental tools to always-on infrastructure across healthcare, software, creative work, and heavy industry. As agents gain autonomy, the real race is shifting toward who controls the platforms orchestrating them.