Amazon and AI startup Anthropic announced an expanded strategic arrangement on April 20, 2026 that deepens their cloud and capital ties. Amazon will invest $5 billion immediately and up to $20 billion more tied to commercial milestones (widely reported as up to $25 billion in new funding). Anthropic, for its part, committed to spending more than $100 billion on Amazon Web Services over the next decade and has secured up to 5 gigawatts of capacity on Amazon’s Trainium AI chips (covering Trainium2 through Trainium4). Companies said nearly 1 GW of Trainium2/3 capacity will come online by year-end and that the full Claude platform will be available within AWS accounts. The deal follows Amazon’s similar, larger package with OpenAI and comes as hyperscalers race to lock major AI labs to their custom silicon. Amazon had previously invested roughly $8 billion in Anthropic; combining prior and potential future payments, total support could approach the low tens of billions.

Apple on April 20, 2026, named longtime hardware executive John Ternus as its next chief executive, handing over day-to-day control to the 50-something engineer on Sept. 1 and elevating Tim Cook to executive chairman. Ternus, a 25-year Apple veteran who has overseen development across iPhone, iPad, Mac, Apple Watch and AirPods, has led recent hardware revamps including the iPhone Air and the shift to Apple silicon. The move accompanies an internal hardware leadership reshuffle — Johny Srouji was named chief hardware officer — and comes as Apple confronts mounting competitive pressure in artificial intelligence from firms such as Nvidia, Google and Meta. The company has begun using Google’s Gemini to bolster Siri and faces investor scrutiny after losing top market-value position to AI chipmakers. Apple shares dipped modestly after the announcement. The board said the planned, unanimous succession reflects long-term planning as Apple, now roughly 50 years old, seeks to accelerate product-led responses to the generational shift driven by AI and mixed-reality devices.

Amazon on April 20-21 agreed to pump $5 billion into AI developer Anthropic and made contingent commitments that could raise its total new funding to $25 billion, part of an expanded multi-year infrastructure pact. Anthropic, which had previously taken about $8 billion from Amazon, has pledged to spend more than $100 billion on Amazon Web Services over the next decade and secured up to 5 gigawatts of capacity on Amazon’s Trainium custom AI chips (covering Trainium2 through future Trainium4 generations). The companies said Anthropic will bring nearly 1 GW of Trainium2/3 capacity online by year-end and will make the full Claude platform available directly inside AWS. Amazon’s investment is structured with an initial $5 billion and up to $20 billion tied to commercial milestones; the move follows Amazon’s large investment and cloud deal with OpenAI earlier this year. Anthropic said surging demand strained its infrastructure and that the AWS commitments are intended to expand training and inference capacity for its Claude models. Reports also cite Anthropic’s accelerated revenue growth and a recent valuation in the hundreds of billions of dollars.

Meta has begun limited tests of a paid WhatsApp tier called “WhatsApp Plus,” rolling out to select Android beta users from April 20–21, 2026, with iOS support promised later. The optional subscription is largely cosmetic and organisational: premium stickers and animated packs, 18 chat themes and alternative app icons, custom ringtones, more granular chat-list controls, and an expanded pin limit (up to 20 chats versus the free tier’s three). Reported regional pricing varies (e.g., €2.49 in Europe, PKR 229 in Pakistan, MX$29 in Mexico; some outlets cite roughly $2.99/month), and Meta may offer a one-month free trial. WhatsApp’s core messaging, voice/video calls and end-to-end encryption remain free, and there is no indication paid plans remove Status ads. The test follows the recent Instagram Plus trial and forms part of Meta’s broader push to add consumer subscriptions across its apps as it seeks to diversify revenue while continuing heavy investment in AI infrastructure.

Adobe on April 20 unveiled CX Enterprise, a new suite of AI agents designed to automate and personalise digital marketing, customer engagement and sales workflows for large organisations. The platform — which includes an autonomous worker called CX Enterprise Coworker — aims to let enterprises embed AI into existing processes and orchestrate agents across ecosystems. Adobe said it has partnered with major cloud and model providers including Amazon Web Services, Microsoft, Google Cloud, Nvidia, Anthropic, OpenAI and IBM to ensure cross-platform compatibility. The announcement, made at Adobe Summit, lifted the stock roughly 2–2.5% intraday after a year-to-date drop of about 30%. Adobe remains under investor scrutiny as competitors such as OpenAI, Anthropic and design-focused entrants press into image-generation and creative tools; reports also note OpenAI preparing a new image model. Adobe’s market metrics cited by analysts show a P/E around 14–15 and a market value near $100 billion. The company is also navigating leadership change after CEO Shantanu Narayen signalled plans to step down after 18 years, adding an extra layer of investor focus on whether CX Enterprise can convert into durable enterprise revenue.

Google said on April 20 it is expanding its Gemini in Chrome assistant to seven additional markets — Australia, Indonesia, Japan, the Philippines, Singapore, South Korea and Vietnam. The feature appears in Chrome’s sidebar and is available on desktop and iOS in all listed countries except Japan, where iOS support is not yet offered. Gemini in Chrome connects to Google services through Personal Intelligence, allowing users to draft and send emails via Gmail, add events to Calendar, check locations in Maps and surface photos from Google Photos. The sidebar also provides access to Google’s Nano Banana 2 image generator. The rollout follows earlier launches in the U.S. (January) and expansions to India, Canada and New Zealand in March. Google noted that its more agentic capability — which can act across the browser to complete tasks — remains in testing and is restricted to AI Pro and AI Ultra paid plans for U.S. users.