Quick wins for a faster PC:
Repair Windows errors before they cause bigger problemsFix Now →Scan for outdated or missing drivers - takes under a minuteDriver Scan →Clear out junk files and repair common Windows errorsFree Scan →Amazon announced on May 14, 2024, that Adam Selipsky would step down as chief executive of Amazon Web Services on June 3. Matt Garman, then AWS’s senior vice president for sales, marketing and global services, succeeded him. Amazon presented the change as planned succession: Selipsky said he wanted time with his family, to recharge and consider his next challenge, and the company did not announce a dismissal.
The timing nevertheless mattered. AWS was entering a cloud market reshaped by generative AI, huge infrastructure commitments and highly visible competition from Microsoft and Google. Garman’s promotion therefore represented continuity of AWS’s operating culture, combined with a more execution- and customer-focused mandate to turn that infrastructure base into faster AI adoption.
What happened to Adam Selipsky at AWS?
Amazon’s May 14, 2024 announcement said Selipsky would leave the AWS CEO role effective June 3. Garman took over that day. Selipsky was not described as fired, and Amazon did not name a successor employer or say he was retiring permanently. Its statement said he planned to spend time with family, recharge and reflect before deciding on his next step.
Andy Jassy said succession planning had been part of the arrangement when Selipsky became AWS CEO. The announcement also praised Selipsky’s results and said AWS was in a strong position with a prepared leadership team. That is evidence of a planned handoff, although it does not rule out strategic or performance pressure surrounding the timing.
#1 Best Overall
Why the timing was significant
Selipsky left as cloud priorities were shifting from pandemic-era migration toward generative-AI models, accelerated computing, networking, data centers and custom chips. Microsoft had prominent OpenAI ties, while Google promoted its own models and AI infrastructure. Analysts and some customers viewed AWS as less visible or less aggressive in the first phase of the generative-AI cycle.
“Behind in AI” needs precision. AWS already offered machine-learning services, Bedrock, SageMaker, custom silicon and partnerships. The criticism generally concerned mindshare, model partnerships, product timing or developer excitement—not an absence of AI capability or proof that AWS had lost its overall cloud position. Forrester’s contemporaneous analysis captures that strategic debate: https://www.forrester.com/blogs/with-selipsky-out-whats-next-for-aws/.
Selipsky’s record at AWS
Selipsky returned to Amazon after previously leading AWS sales, marketing and support. He became AWS CEO in 2021 when Jassy became Amazon’s chief executive. During his tenure, AWS continued adding infrastructure and enterprise services and launched generative-AI products including Amazon Bedrock and Amazon Q.
Jassy said AWS had passed a $100 billion annualized revenue run rate in the quarter before the transition. Amazon’s 2024 shareholder letter later reported $108 billion of AWS revenue for 2024, up 19% year over year: https://www.aboutamazon.com/news/company-news/amazon-ceo-andy-jassy-2024-letter-to-shareholders. Growth had slowed from pandemic peaks, however, while customers focused on optimizing cloud spending and the industry redirected attention toward AI infrastructure.
Do these 3 things before closing this tab:
1Repair Windows errors before they cause bigger problems2Fix the driver behind crashes, sound loss and screen glitches3Clear out junk files and repair common Windows errorsWho is Matt Garman?
Garman joined Amazon as an MBA intern in 2005 and became a full-time employee in 2006. He was among AWS’s early product managers and subsequently worked across product management, engineering, operations, sales, marketing and global services. Immediately before becoming CEO, he was senior vice president for AWS sales, marketing and global services.
Amazon’s official biography shows why the appointment was not an outsider’s restructuring. Garman had deep knowledge of AWS products and operations, plus direct experience with customers and commercial execution. That combination positioned him to preserve reliability and service breadth while making AI adoption easier to sell, deploy and scale.
Garman’s mandate
The new CEO inherited a business with several simultaneous objectives:
- Sustain AWS’s broad cloud business while growth reaccelerated.
- Make generative-AI workloads economically viable for enterprises.
- Expand Trainium and Graviton custom chips alongside third-party accelerators.
- Use Bedrock and SageMaker to give customers model choice and production tooling.
- Bring inference close to customers’ existing data and applications.
- Retain customers facing multicloud pressure without sacrificing security, reliability or operating discipline.
Amazon’s strategy emphasizes that AI demand also drives conventional cloud consumption: storage, databases, networking, security, compute and data-management services. Its discussion of that approach appears in Amazon’s AWS AI strategy overview. The practical challenge is converting broad technical capability into repeatable production workloads rather than short-lived experiments.
Crashes, No Sound, or Screen Glitches?
Random freezes, missing sound and display glitches usually trace back to one bad driver. Find and replace yours safely.Free scan · under a minutePC Slower Than It Used to Be?
A free scan shows the junk files, broken settings and background clutter dragging Windows down - then fixes them in one click.Free scan · Windows 10 & 11Rank #3
Was this routine succession or a strategic correction?
| Evidence for planned succession | Evidence of strategic pressure |
|---|---|
| Jassy said the next generation of AWS leaders had been prepared. | Generative AI shifted customer and investor attention toward models, accelerators and infrastructure. |
| The announcement praised Selipsky’s results and AWS’s position. | Microsoft’s OpenAI relationship and Google’s model visibility raised competitive expectations. |
| Selipsky’s message described time with family and a next chapter. | Analysts questioned whether AWS had moved quickly enough in the earliest generative-AI cycle. |
| Garman was a long-serving internal AWS executive. | The new CEO had to accelerate execution while capital spending and capacity requirements surged. |
The public record supports both facts: this was a planned internal transition that occurred amid real strategic pressure. It does not support calling Selipsky’s departure a firing, nor claiming Garman was appointed solely to repair an AI failure.
Did the “future is bright” claim hold up?
Later figures support the broad optimism, while leaving causation unresolved:
| Measure | Reported result | Source and qualification |
|---|---|---|
| AWS 2024 revenue | $108 billion, up 19% year over year | Amazon’s 2024 shareholder letter |
| AWS growth in Q1 2026 | 28% year over year | Amazon Q1 2026 results |
| AWS Q2 2026 sales | $42.2 billion, up 36.7% year over year | Amazon Q2 2026 results; company-reported growth, fastest in 18 quarters |
| AWS Q2 2026 annualized run rate | $169 billion | Amazon-reported annualization |
| AWS Q2 2026 operating income | $16.6 billion versus $10.2 billion in Q2 2025 | Amazon-reported quarterly figures |
| AWS AI business | More than $25 billion annual revenue run rate, with triple-digit year-over-year growth | Amazon-reported; includes the AI business, not total AWS |
| AWS chips business | More than $25 billion annual revenue run rate | Amazon-reported in its chips history |
These results show strong momentum under Garman and support the direction of AWS’s infrastructure and AI strategy. They do not prove that changing CEOs caused the acceleration. Results also reflect industry-wide AI demand, customer migrations, chip availability, pricing and product changes, Amazon’s investment levels and broader economic conditions. The reported figures cover all AWS, not only AI.
Risks behind the growth
- Capital intensity: Data centers, power, networking and accelerators require enormous spending before revenue arrives.
- Supply constraints: GPU and custom-chip availability can limit how quickly customers deploy models.
- Demand normalization: Customers may reduce cloud usage after initial AI experiments or shift workloads among providers.
- Platform economics: AWS must balance Trainium and other proprietary silicon with Nvidia and competing accelerators.
- Customer bargaining power: Large model developers and enterprises can negotiate aggressively or use multiple clouds.
- Complexity: Service-specific pricing, data-transfer charges and usage-based billing can make AWS difficult for smaller teams without FinOps controls.
Amazon reported that trailing-twelve-month free cash flow became a $7.6 billion outflow through Q2 2026, driven primarily by higher property-and-equipment spending tied to AI investment: https://www.aboutamazon.com/news/company-news/amazon-earnings-q2-2026-report. Faster revenue growth therefore does not automatically mean stronger near-term cash generation.
What the transition means for AWS customers
Expect continuity, not a wholesale product reset
Garman’s internal promotion makes continuity in core compute, storage, databases, networking, security and support more likely than a sudden reorganization. Customers should not assume, however, that “no change” means unchanged pricing, sales priorities, regional capacity or AI-service availability.
AI capacity and tooling may receive greater emphasis
Customers can expect continued investment in Bedrock, SageMaker, model access, custom chips and inference infrastructure. The relevant test is workload performance: price, latency, reliability, portability and integration with existing data—not executive messaging alone.
Keep alternatives and exit costs visible
AWS remains a natural fit for organizations already invested in its services or needing broad infrastructure coverage. Azure may merit comparison for Microsoft-centric enterprises, Google Cloud for some analytics, Kubernetes or AI workloads, and Oracle Cloud Infrastructure for Oracle-heavy estates. No provider is universally cheapest or technically superior without a defined region, traffic profile, workload and commitment period. Official starting points are AWS pricing, Azure pricing, Google Cloud pricing and Oracle’s cost estimator.
For a proof of concept, compare accelerator availability, total cost including data transfer, model portability, inference latency, security and identity integration, committed-use discounts, operational support, exit costs and the internal expertise required.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
Implications for investors, employees and competitors
Investors should separate AWS’s operating performance from the question of who caused it. The figures indicate a valuable growth engine, but AI-related capital expenditure can pressure free cash flow and returns. Employees saw an internal promotion that preserved institutional knowledge while raising expectations for faster execution. Competitors face an AWS that can combine a large installed base with custom silicon, multiple models and enterprise distribution rather than competing only through a single chatbot.
The most defensible interpretation is therefore continuity with a sharper operating mandate. Selipsky’s departure was publicly framed as planned succession, and Garman was chosen because he knew AWS deeply and had spent years close to customers and commercial execution. Subsequent growth makes the “bright future” judgment look credible, but it does not establish that the leadership change itself caused the result or that every AWS strategic bet will succeed.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
