OpenAI has put conversations and Projects back in its redesigned ChatGPT desktop app and enabled cloud Work threads to move across devices, correcting the launch's biggest usability failures without merging local Work or Codex histories.
Dave Treadwell will replace departing AWS veteran Dave Brown as head of Compute and Machine Learning Services on August 1, moving one of the cloud unit’s largest internal customers into a supplier role as Amazon sharply increases infrastructure investment.
Amazon is moving the executive responsible for the technology foundation of its online stores into one of AWS’s most consequential supplier roles. The appointment gives Dave Treadwell oversight of compute and machine-learning services while Amazon is adding infrastructure at a much faster rate, but it does not by itself show what he will change.
Treadwell will become senior vice president of AWS Compute and Machine Learning Services on August 1, according to employee messages published in Amazon’s announcement. AWS CEO Matt Garman said Brown had decided to take another job outside Amazon. Brown will stay through the end of July to help with the transition; Amazon has not identified his next employer.
That title matters. Treadwell will lead a major AWS organization, not replace Garman as chief executive of the cloud business. Brown’s remit included the EC2 computing service and machine-learning products such as Bedrock and SageMaker, an account of the succession said.
Brown joined AWS in Cape Town in 2007 as a developer working on EC2. His departure follows his promotion to senior vice president in April and comes after he joined Amazon CEO Andy Jassy’s senior leadership team in 2023. That group had 28 members including Brown at the time of the report. Treadwell is also a member.
The timing makes the exit notable, but the available evidence does not establish a setback inside Brown’s organization. Brown told employees that it felt like the right time to start a new chapter, while Garman said AWS had never been in a stronger position. Neither statement explains the choice beyond Brown accepting an outside role.
Treadwell has led eCommerce Foundation, the organization behind the technology platform for Amazon’s stores. Garman described him as one of AWS’s largest and most vocal internal customers and said he had run Amazon’s technology platform at comparable scale for nearly a decade.
That experience is the clearest rationale Amazon has offered for the move. It also needs qualification. In a company profile, Amazon said Treadwell led site availability for Prime Day and developed engineering processes intended to reduce the stores business’s infrastructure-cost trajectory. Treadwell called Prime Day 2026 the event’s cleanest and smoothest edition and said those cost processes remained in use. Those are Amazon’s and Treadwell’s claims; the company supplied no independent performance figures for either achievement.
His background extends beyond Amazon. Treadwell graduated from Princeton in 1989 with an electrical engineering degree and spent 27 years at Microsoft, where his work included Windows, Xbox and the .NET software framework, according to an account of the transition. He joined Amazon in 2016.
The move therefore installs an experienced application operator and internal buyer, not an outside turnaround executive. It does not prove that the priorities of Amazon’s retail platform transfer directly to external AWS customers, whose workloads, contractual terms and alternatives vary widely.
AWS is expanding quickly. In the first quarter of 2026, segment sales rose 28% from a year earlier to $37.59 billion, while operating income increased from $11.55 billion to $14.16 billion, Amazon said in its quarterly filing. The company attributed the sales increase mainly to greater customer usage, partially offset by pricing changes driven mainly by long-term customer contracts.
The same filing shows the infrastructure burden behind that growth. AWS’s net additions to property and equipment increased from $20.46 billion in the first quarter of 2025 to $41.52 billion a year later. Depreciation and amortization allocated to the segment rose from $4.39 billion to $7.28 billion.
Some demand is already under long-term contract. Amazon disclosed about $364 billion of performance obligations, primarily related to AWS, with a weighted-average remaining life of 5.5 years; revenue depends on customer usage and Amazon meeting those obligations. It also said AWS and OpenAI expanded an existing $38 billion multi-year commercial commitment by $100 billion over eight years, including obligations tied to the performance of AWS chips. Those commitments make usage, capacity delivery and chip performance central to whether contracted demand becomes revenue.
Amazon’s broader cash capital expenditure reached $43.2 billion for the quarter, up from $24.3 billion. The company said the spending primarily covered technology infrastructure—most of it supporting AWS growth—and additional fulfillment capacity. That comparison is company-wide, not a budget assigned to Treadwell’s organization.
Nor is the financing behind Amazon’s expansion his disclosed mandate. Amazon raised $53.44 billion of long-term debt during the quarter and said its March note offerings were for general corporate purposes. The filing does not earmark those proceeds for AWS or give the incoming compute leader control over them.
Infrastructure constraints cut in both directions. Amazon warned that inaccurate demand forecasts can leave it with excess or insufficient data-center capacity, higher costs or service interruptions. It also identified reliance on a limited group of semiconductor suppliers, including suppliers of graphics processors for AI infrastructure, as a risk. Treadwell’s retail experience may be relevant to capacity discipline, but AWS must make those choices for external demand rather than a single internal application estate.
Competition further limits the claim that this is a rescue assignment. An independent market estimate put worldwide cloud-infrastructure spending at $128.6 billion in the first quarter, up 35% year over year. It estimated shares of 28% for Amazon, 21% for Microsoft and 14% for Google, while finding that Microsoft and Google were growing substantially faster. The measured market covers infrastructure as a service, platform as a service and hosted private cloud, so its growth and share figures are not directly comparable with Amazon’s AWS segment results.
The alternatives are also broader than those three companies. The same analysis said fast-growing second-tier providers included CoreWeave, Oracle, Crusoe and Nebius. It counted five “neocloud” companies among the top 30 providers and estimated that neoclouds as a group represented 5% of cloud-infrastructure revenue. Treadwell inherits a leader in a rapidly expanding market, but one facing faster-growing hyperscale rivals and AI-focused capacity specialists.
The immediate succession is settled; the operating mandate is not. Amazon has not announced a reorganization under Treadwell, a different product roadmap, revised pricing priorities or a new method for allocating compute capacity. It has also not said who will assume his responsibilities in eCommerce Foundation.
The first meaningful evidence will be specific decisions rather than leadership biography: changes to the compute and machine-learning organization, disclosed shifts in product or capacity priorities, and customer-facing changes to price, reliability or availability. Financially, AWS usage growth, operating income and the cost absorbed as new infrastructure enters service will show whether the buildout is becoming productive, although those segment-wide outcomes will remain broader than Treadwell’s remit.
Until those signals appear, the defensible conclusion is limited. Amazon has chosen a technically experienced internal customer to succeed an early EC2 builder at a capital-intensive moment. Whether that buyer’s perspective changes AWS economics or execution is the question the appointment creates, not an outcome the announcement establishes.
Meta reportedly plans to put departing AWS compute executive Dave Brown to work on its data-center buildout, adding hyperscale operating experience while leaving any customer-facing cloud business conditional and undefined.
China has paired a five-year AI training offer for developing countries with cooperation centers, a weather-warning rollout and a new 29-country organization. The package gives Beijing a platform for influence, but no budget, selection rules or delivery timetable has been published.
RoboTTT extends a robot policy’s visuomotor context to 8,000 timesteps and raises its average rubric score from 42% to 79%, but the evidence comes from three author-run assembly tasks and the longest task was completed in only two of ten trials.
Zhipu reportedly reached $1 billion in annual recurring revenue in July, roughly four times a March estimate, but the unconfirmed run rate is not annual sales and still sits far ahead of recognized cloud revenue while margins remain thin.
Nebius has arranged its first senior secured facility against an operating GPU deployment and one customer’s cash flows. The deal adds project-level debt to its expansion toolkit, but an unnamed customer and a larger rival financing limit what it proves about the rest of Nebius’s backlog.
Cursor says automatically running a repository-root git.exe on Windows does not meet its criteria for patching, while the researcher calls it an untrusted-search-path defect and separate research shows the same weakness across several competing AI coding tools.
Huawei publicly displayed a 16-cabinet Atlas 950 configuration rated at 1 EFLOPS in FP8, providing tangible evidence of its system-scale AI strategy while leaving price, power use and sustained workload performance undisclosed ahead of the full system's planned fourth-quarter release.
CIA Director John Ratcliffe said US intelligence is consistent with an estimate that Russian recruits last 20 to 30 minutes on Ukraine’s battlefield, but the public trail leads to an unsourced claim about assault troops and does not establish a representative average.
UK testing places leading open-weight models four to seven months behind selected closed-model cyber results, yet longer attack chains, U.S. benchmarks and mixed cost comparisons show why that interval is a warning signal rather than a universal capability clock.
TSMC reached the top of its second-quarter revenue guidance and raised its 2026 outlook, but a one-off investment gain boosted profit growth while the company committed more capital to 2-nanometer production, advanced packaging and an undated Arizona expansion.
Moonshot AI's Kimi K3 added to a global technology selloff with near-frontier performance, but unreleased weights, mixed cost comparisons and a recommended 64-accelerator deployment leave its effect on chip demand unresolved.
Moonshot AI has made Kimi K3 available through its apps and API, pairing a 2.8-trillion-parameter architecture with early frontier-level results, but the model's open-weight claim cannot be tested until its weights and technical report arrive.
Two binding EU decisions require Google to give rival AI services comparable access to 11 Android features and offer eligible search competitors a restricted, anonymized dataset, but phased deadlines, certification, pricing and privacy safeguards leave the competitive effect unproven.
Databricks has signed a term sheet for a Coatue-led financing at a $188 billion valuation, while unidentified sources put the round at $3 billion. The proposed capital would deepen its push into AI governance, data agents and operational databases, but the transaction remains open and the company supplied no new operating figures.
Netflix says generative AI workflows were used on roughly 300 titles in 2026, but its only quantified example tied to that disclosure covers 17 minutes and does not establish how much finished material, spending or labor changed across the slate.
Gemini 3.5 Pro missed its expected June rollout. An anonymous-source account says a late-June data change fell short of Google's coding goals, but Google has confirmed only partner testing—not the reported cause, a new date, or public results and pricing.
Google is bringing Gemini Omni editing and a reusable face-and-voice avatar to Vids, but the sharper distinction is account-level identity across Vids and Gemini rather than a new category of AI video; Vids already offered Veo generation and customizable avatars, while specialist rivals already sell digital presenters.
Intel plans to adopt Gemini Enterprise across engineering, supply-chain and corporate work while extending chip-simulation capacity into Google Cloud, but the companies supplied no rollout schedule, contract value or performance baseline.
The Navy has approved an immediate department-wide framework for turning data into operational effects, but the public rollout leaves budgets, owners, deadlines, performance measures and assurance rules to a promised implementation roadmap.