AI Works

XPeng's chief executive took direct leadership of the company's robotics division as the Chinese EV maker said humanoid mass production was approaching. The move signals that senior executives at major Chinese tech-and-hardware companies now treat embodied AI as core business rather than a side project — a pattern also visible at BYD and other Chinese industrial groups expanding into physical AI.Reuters →

Figure AI reported producing hundreds of Figure 03 units and ramping manufacturing throughput. The announcement marked a shift from proof-of-concept demos toward something closer to a volume hardware operation, making Figure one of the first US humanoid robotics startups to publish production numbers rather than capability milestones alone.Figure AI →

Figure AI signed a commercial agreement to deploy humanoid robots in Catalyst Brands' logistics and distribution network, one of the first announced commercial contracts for humanoid robots in a live retail supply chain. The deal is notable less for its scale than for what it represents: a shift from demo deployments in controlled environments toward contractual commitments to run humanoids in real logistics operations.Figure AI →

NVIDIA's CEO confirmed the company is working with LG on humanoid robotics and data-centre infrastructure. The partnership adds a major Korean industrial conglomerate to the network of hardware partners building on NVIDIA's physical-AI stack, which spans simulation (Isaac Sim), model development (Cosmos, GR00T), and compute. LG's involvement points to humanoid manufacturing capacity and consumer-facing deployment scenarios beyond the factory-floor use cases that have dominated the category.Reuters →

Meta rolled out Business Agent — an AI assistant businesses can deploy across WhatsApp, Messenger, Instagram, and Meta Business Suite to handle customer queries, generate content, and manage interactions at scale. The deployment pushes AI agents into the mainstream messaging channels used by billions of people, making it one of the largest-reach agent launches to date in terms of potential end-user exposure.Meta Newsroom →

Microsoft made Discovery generally available on Azure as an agentic AI platform for scientific and engineering research — supporting simulation, hypothesis generation, materials analysis, and autonomous lab workflows. The GA release moves an experimental product into a billable Azure service, signalling Microsoft's intent to position agentic tools not just for software productivity but for physical-science and engineering R&D pipelines.Azure blog →

At Build 2026, Microsoft announced a set of enterprise AI products under the IQ family — Microsoft IQ, Work IQ, Fabric IQ, and Foundry IQ — providing agentic context, knowledge indexing, and work-automation APIs for organisations running on Microsoft infrastructure. The announcements position Microsoft 365 and Azure as the coordination layer for enterprise agents that need access to organisational knowledge and permissions without bypassing security boundaries.Microsoft blog →

NVIDIA released a set of physical-AI agent skills covering simulation, synthetic data generation, policy training, and evaluation — the tooling layer that lets robotics teams build, test, and iterate on robot behaviour without requiring physical hardware at every step. The release accompanies Cosmos 3 and Isaac GR00T, rounding out NVIDIA's physical-AI stack from foundation model to development workflow.NVIDIA blog →

Coinbase released Coinbase for Agents, a product suite giving AI agents direct access to crypto payment rails, trading functions, and wallet management. The launch is one of the clearest signs so far that AI agents are being given financial execution capabilities — moving from planning and content tasks into actions with real monetary consequences. Developers can now wire agents to spend, receive, and move funds without human sign-off on each transaction.Coinbase blog →

OpenAI added direct email-sending to ChatGPT writing blocks, letting users compose and dispatch messages from within a conversation rather than copying output to a separate mail client. The feature is a small but concrete step into action-taking territory — ChatGPT was previously a content generator; it can now initiate external communication on the user's behalf without leaving the interface.ChatGPT on X →

Google announced the next evolution of the Gemini app, adding proactive suggestions, more persistent context, and assistant-like capabilities that can act across apps and services. The changes push Gemini closer to the ambient-assistant model that Apple and OpenAI are also pursuing — an AI that anticipates needs rather than waiting to be prompted.Google blog →

Google updated NotebookLM with improved research and reasoning capabilities, moving the product further from a summarisation tool toward a structured research assistant that can plan enquiries, cross-reference sources, and synthesise findings. The upgrade signals Google's intent to make NotebookLM a direct competitor to dedicated research tools and agentic search products rather than a document-reading companion.Google blog →

Google announced live speech translation as a feature of Gemini, enabling spoken conversations to be translated in near real time. The capability targets travel, cross-language meetings, and accessibility use cases where current translator apps require a stop-start exchange rather than fluid back-and-forth. Live translation has been a persistent aspiration for AI assistants; embedding it in a general-purpose model rather than a dedicated app changes how easily it can be accessed.Google blog →

Apple introduced a suite of AI developer tools at WWDC 2026 — Foundation Models for on-device inference, Core AI for app integration, and agentic capabilities inside Xcode for code generation and debugging. The release opens Apple's private-cloud and on-device AI stack to third-party developers, making it possible to build intelligence features that use Apple hardware without sending data to external providers.Apple Newsroom →

Apple used WWDC 2026 to announce a substantially rebuilt Siri — now called Siri AI — with deep personal context, app actions, and on-screen awareness, alongside a broad Apple Intelligence update bringing AI into Photos, writing tools, and system-level tasks. The private-cloud architecture keeps most processing on-device or within Apple's encrypted infrastructure. For many consumers, Siri AI will be their first hands-on experience of a persistent, contextual AI agent built into a device they already own.Apple Newsroom →

Microsoft introduced a redesigned Microsoft 365 Copilot with a cleaner prompt experience, inline formatting and tighter integration across Word, Excel, PowerPoint and Outlook. The update reworked the assistant's interface and performance across the Office surface. The redesign reflected Microsoft's effort to make Copilot a consistent layer across its productivity apps rather than a bolt-on panel, as competition for the in-document AI assistant intensified.Microsoft 365 blog

Microsoft announced general availability of computer-using agents in Copilot Studio, alongside a new workflows experience and real-time voice capabilities. The agents can operate software interfaces directly to complete multi-step tasks, moving Microsoft further into enterprise automation. The GA milestone arguably mattered more for businesses than the parallel Copilot redesign, signalling that agent execution — not just chat assistance — is now a shipping enterprise feature rather than a preview.Microsoft Copilot blog

ElevenLabs launched Music v2, a music-generation model with improved vocals, instrumentation and arrangement, multilingual support and licensing aimed at commercial use. The model generates full songs from a text prompt, extending the company's reach from voice into produced music. The release deepened competition in AI music alongside Suno and Google's Lyria line, and pushed the licensing question — who owns and can monetise AI-generated songs — further into the mainstream.ElevenLabs announcement

ElevenLabs released Dubbing v2, an AI dubbing model that preserves a speaker's emotion, tone, delivery and performance when translating speech across languages. The system targets film, video and creator localisation where flat machine dubbing has historically fallen short. Performance-preserving dubbing lowers the barrier to releasing video content in many languages at once, with obvious implications for the human voice-acting and localisation industries.ElevenLabs announcement

Perplexity launched an integration that brings its answer-engine and agent-style workflow into Word, Excel, PowerPoint and Outlook, distributed through the Microsoft marketplace. The move placed a third-party AI assistant directly alongside Microsoft's own Copilot inside Office. The integration showed how the productivity suite is becoming contested territory, with independent AI players seeking distribution inside the apps where knowledge workers already spend their day.Microsoft Marketplace listing

Leonardo.ai rolled out a workflow that turns a single 2D image into a 3D model exported as a .glb asset, aimed at creators, product designers and game developers. The tool converts AI-generated or uploaded images into manipulable 3D meshes without manual modelling. Image-to-3D pipelines continued to mature across the industry in 2026, shortening the path from concept art to usable game and product assets and lowering the cost of 3D content for small studios.Leonardo.ai guide

At Google I/O 2026, Google unveiled a set of agentic consumer features built on the Gemini 3.5 family. Gemini Spark is a 24/7 always-on personal agent running on Google's servers and operating across Workspace, with Daily Brief producing a personalised digest and action list. Universal Cart adds a unified shopping experience that lets Gemini buy across the web and partner retailers; AI Inbox brings agentic management to Gmail; Keep Live and Docs Live add voice-driven organisation and editing; Ask YouTube enables conversational video search with timestamps. Google highlighted MCP / Anthropic-protocol support for connecting external apps, and pointed to its 900M+ Gemini monthly-active-user base as the distribution channel. Taken together, the announcements re-frame Google's product surface — Search, Gmail, YouTube, Android, Chrome — as an agentic platform rather than a set of separate tools, and lean on the same Gemini 3.5 / Omni model stack Google announced under the Road to AGI banner.Google I/O 2026 announcements

Figure AI ran a public 24/7 livestream of its humanoids sorting packages on a warehouse-style line for more than 34 hours straight. The stream was meant to demonstrate sustained, unscripted operation rather than the short, polished clips that usually dominate humanoid demos — including stumbles, recoveries and slow stretches where the bots queued for parts. Coverage was mixed. Supporters pointed to it as the first long-window evidence of a humanoid keeping up with real industrial throughput; sceptics noted the bots' pace, the controlled environment and Figure's recent $39 B post-money funding round as reasons to read the stream as part marketing, part proof.ZeroHedge coverage of the livestream

Rivian shipped a software update that adds an in-car AI assistant tuned to the vehicle. It can read incoming messages aloud, answer owner-manual questions in natural language, surface trip and charging data, and toggle settings such as climate or drive modes by voice — without forcing the driver to navigate touchscreen menus. Rivian framed the assistant as a step away from generic smartphone assistants embedded into the dashboard, towards an agent that knows the specific vehicle's state. The update rolls out over the air to existing R1S and R1T owners on the latest software branch.Rivian software update

Meta added an Incognito mode to WhatsApp Meta AI conversations that drops chat history and prevents the session being used for personalisation or model improvement. In parallel, Meta widened access to Muse Spark, the most capable model from its Superintelligence Labs unit, moving it from a small preview to a broader rollout inside Meta apps. The two changes line up Meta's consumer AI story: a privacy mode for sensitive conversations, and a stronger underlying model so users have a reason to pick Meta AI over the competing assistants embedded in their phones.Meta announcement

Google rolled out a deeper integration of Gemini into Android, branded Gemini Intelligence, that places multimodal context awareness across the home screen, notifications and apps rather than gating it behind a single assistant surface. In the same announcement, Google introduced Googlebook — its AI-first successor to the Chromebook line, designed around always-available agents and cross-device handoff. The pair of launches positions Android and a refreshed laptop tier as the everyday delivery vehicle for Gemini, set against an iPhone where Apple still depends on third-party model providers for the heavier AI features.Gemini Intelligence on Android

Anthropic split its enterprise push into two vertical bundles. Claude for Legal targets law firms with workflows for contract review, discovery, deposition summaries and case-law research, with a dedicated retention policy intended to meet legal-industry confidentiality expectations. A separate Claude for Small Business package ships ready-made agents — bookkeeping helper, customer-support triage, marketing copy — designed to be deployed without an in-house engineering team. The two bundles continue Anthropic's pattern of going industry-by-industry rather than launching a single horizontal enterprise product, betting that pre-packaged agent recipes will land faster than generic platforms.Claude for Legal post

Anthropic added an Agent View to Claude Code that consolidates several simultaneous agent runs into a single window, replacing the practice of opening a terminal tab per agent. Users can see each agent's current task, intervene mid-run, and route follow-up prompts to a specific worker without context-switching across windows. The change is aimed at the heavy agent workflows that have become common over the past six months — splitting a feature across three or four parallel Claude Code sessions to refactor, test and document at the same time. The new view makes that pattern less chaotic and lowers the barrier for less technical users.Anthropic announcement

OpenAI extended Codex with a mobile workflow that lets developers steer an active coding session on their desktop from a phone — browsing repository files, reviewing diffs and queuing further work while away from their main machine. The interface targets long-running agent loops that previously required staying in front of a laptop to monitor and unblock. The launch follows OpenAI's pattern of making Codex a multi-surface product rather than a single CLI: the agent now spans terminal, IDE, web and a remote-control mobile companion. Heavy coders described the change as a way to keep agent runs productive during travel and meetings.OpenAI release page

Japan Airlines and GMO Internet Group launched a trial in May 2026 to use Unitree G1 humanoid robots alongside ground crews at Tokyo's Haneda Airport, moving baggage and cargo on the tarmac. The 35 kg, 1.32 m G1 has 23 degrees of freedom, 3D LiDAR and a depth camera, and is trained in Nvidia Isaac Sim before deployment. The programme runs through 2028. JAL frames the rollout as a response to a shrinking domestic workforce and record inbound tourism — 42.7 million visitors in 2025 and 7 million in the first two months of 2026. Safety-critical ground tasks remain under human control; the airline plans to expand the rollout if early stages prove out.Japan Airlines press release

Anthropic launched Claude Design, a visual-work product that lets users describe a design brief and receive first-pass layouts, prototypes, slides, and one-pagers, then refine them through conversation, inline comments, direct edits, or custom sliders. The tool is powered by Claude Opus 4.7, runs as a research preview for Claude Pro, Max, Team, and Enterprise subscribers, and can read a team's codebase and design files during onboarding to adopt existing colours, typography, and components. Output can be shared via internal URL, saved as a folder, or exported to Canva, PDF, PPTX, or standalone HTML.Anthropic ↗

Engineer Tony Allevato published a long-form first-person account of living with a consumer AI robot, paired with a companion post on reviving an older unit. The piece covers practical limits, family adjustments and behavioural quirks of consumer-grade home robotics in 2026, with a particular focus on how children interact with a chatty home robot. The post drew 28 comments on Hacker News, where readers raised confirmation-bias concerns about kids learning from a perpetually agreeable robot — a recurring theme as Figure, 1X and Unitree push consumer humanoids. It is a useful primary-source counterweight to vendor demo reels.allevato.me ↗

Microsoft released Agent Framework 1.0 — a production-grade SDK with stable APIs, a long-term support commitment, and built-in support for the Model Context Protocol so agents can use any compliant tool out of the box. A new browser-based DevUI visualises agent execution, tool calls and reasoning steps in real time. More than 90% of the Fortune 500 already use Microsoft 365 Copilot, and the framework gives those organisations a supported route to wire custom agents into Azure AI Foundry alongside the boxed Copilot product. The release lines up with Salesforce Agentforce and OpenAI Frontier as the first generation of agent platforms enterprises can buy on standard support contracts rather than research previews.Microsoft announcement

OpenAI detailed the next phase of its enterprise business: a build-and-manage agent stack called Frontier, with named production deployments at Oracle, State Farm and Uber. The company said enterprise customers now generate more than 40% of its revenue and are on track to match consumer revenue by the end of 2026. The disclosure marks the first time OpenAI publicly anchored its enterprise story to specific Fortune 500 customers running agents in production rather than pilot. It also confirms a pattern across the sector — agent platforms are no longer demoware; they are paying line items inside multi-billion-dollar IT budgets.OpenAI announcement

Developer Alex Ziskind and others demonstrated that a cluster of four M4 Pro Mac Minis connected via Thunderbolt 5 delivers 144GB of unified memory and a 128-core GPU for approximately $6,000 — with electricity costs of roughly $15 per month. The setup is capable of running 70B+ parameter models locally. The trend went viral in early 2026, with developers ordering three to twelve units at a time and depleting Apple store inventory across multiple regions. Illustrated a shift in the hobbyist and small-business AI deployment landscape: purpose-built Mac Mini clusters now offer a credible self-hosted alternative to cloud inference APIs, with no per-token cost and full data privacy.Apple Insider

Salesforce announced general availability of Agentforce 360 — its enterprise AI agent platform — after twelve months of staged releases (Agentforce 1, 2, 2dx, 3) and Salesforce running its own operations as Customer Zero. By late 2024 the company had already booked more than 1,000 paid Agentforce deals and described the line as the fastest-growing product in its history. Agentforce 360 ties agents directly into Salesforce's CRM data graph, adds reasoning, governance and interoperability layers, and turns the existing Salesforce platform into infrastructure for autonomous workflows. The launch sets a benchmark for what shipped, paid-for enterprise agent deployment looks like at Fortune 500 scale.Salesforce announcement

OpenAI announced that ChatGPT Enterprise and Codex would be made available to all Samsung Electronics employees in Korea and to Device eXperience staff worldwide, spanning R&D, manufacturing, marketing and corporate functions. OpenAI described it as one of its largest enterprise deployments to date. The scale makes it a marker for how quickly frontier AI tools are being pushed across an entire global workforce.OpenAI →

Meta introduced AI Mode for Facebook, a search-style feature that uses Meta AI to answer questions drawing on public content across Facebook and Instagram, alongside new creative tools for photos, videos and sharing suggestions. The move puts a conversational AI layer directly inside one of the world's largest consumer platforms. It pushes generative AI features in front of billions of everyday users rather than a developer audience.Meta →

NHS England announced Microsoft 365 Copilot access for 505,000 clinicians and support staff, following a trial in which users reported saving around 43 minutes a day on administrative work. The stated aim is to cut paperwork and free up more time for patient care. It is one of the largest public-sector AI deployments to date and a test of whether assistant tools deliver real time savings at scale.Microsoft →

Reuters reported that US restrictions on AI access were prompting European firms to spread their reliance across multiple AI providers rather than depend on a single one. Companies are wary of being exposed if access to models, tokens or hosting changes suddenly. The shift is as much a procurement and resilience decision as a technical one.Reuters →

The Financial Times reported that Garfield AI, described as the UK's first AI-regulated law firm, won its first English court case, in a dispute over unpaid freelancer fees. It is an early example of an AI-driven legal service producing a result inside the court system. The milestone suggests AI legal tools are moving from drafting into actual litigation workflows.FT →