Amazon Continues To Cut 16000 Gone

Amazon has announced plans to cut a further 16,000 roles from its corporate workforce, extending the cost and organisational shake-up that has run alongside a sharp rise in spending on artificial intelligence.

In a message to staff sent on Wednesday January 28 2026, Beth Galetti, Amazon’s senior vice-president for people experience and technology, said the group was making further “organisation changes” aimed at “reducing layers”, increasing ownership and “removing bureaucracy”. The latest round follows an earlier plan, set out in late October 2025, to eliminate 14,000 corporate positions.

Streamlining, again

The new cuts take the total number of announced corporate lay-offs since October to about 30,000. Amazon does not publish a single figure for its corporate headcount, but analysts have estimated it at roughly 350,000, with most staff based in the US.

Galetti said the changes would be spread across the organisation. For most US-based employees affected, the company will offer a 90-day window to apply for other internal roles, after which they will be offered severance, according to the notice.

The message was framed as the continuation of a reorganisation that began last autumn, with some teams completing changes in October and others only finishing now. In other words, Amazon is presenting this as unfinished work rather than a fresh, rolling programme of cuts.

Cost discipline meets an AI arms race

The timing underlines the tension running through big tech: layoffs and leaner management structures, at the same time as capital spending accelerates. Amazon has signalled it expects to invest about $118bn in 2025, with much of that directed toward AI infrastructure, data centres and the computing capacity needed to train and run large models.

That spending is concentrated in Amazon Web Services, the company’s biggest profit engine, and the business most exposed to a shift now under way in cloud computing. The cloud market is no longer only about renting storage and processing power for traditional workloads. It is increasingly about AI-specific capacity, specialist chips, and the ability to deliver models and tools that businesses can deploy at scale.

Amazon remains a dominant player in conventional cloud services, but investors have questioned whether it is moving quickly enough in the AI phase of the market. Rivals, including Microsoft and Google, have pushed aggressively into AI products and the chips used to run them, while Amazon has been trying to build momentum around its own software, in-house silicon and model offerings.

A tougher sell for AWS

One challenge is narrative as much as technology. AWS has long been positioned as the default choice for corporate cloud, yet the AI boom has raised a sharper question for customers: what is distinctive about a provider’s AI stack and why should workloads be built there rather than elsewhere.

Analysts at Raymond James have estimated Amazon will generate about $5bn in AI cloud revenues in 2025, compared with around $17bn for Microsoft. Josh Beck, an analyst at the bank, has argued that Amazon has yet to provide clear, high-profile examples of customers choosing its own software and Trainium chips in a way that shifts perceptions of AWS’s place in the AI race.

The comparison is uncomfortable for Amazon because AWS is not only a business line but the financial foundation that funds its wider ambitions. Any sense that AWS is falling behind in the next computing cycle has consequences for investor confidence, even if the core cloud franchise remains sizeable.

Management changes and chip ambitions

The job cuts also land shortly after a senior management reshuffle. Amazon has been adjusting leadership across its AI efforts as it tries to speed up execution and close gaps with competitors on advanced chips and large language models.

The company has invested in Anthropic, the AI start-up behind Claude, and has been expanding its own model offerings. It has also been promoting its in-house chips, including Trainium for training models and Inferentia for running them. However, those efforts have not generated the same external attention as Nvidia’s latest systems or rival cloud providers’ flagship AI launches.

Amazon’s most recent chip line was released late last year, with the group presenting performance and cost advantages for certain workloads. Even so, the market is centred on a small number of reference points, including Nvidia’s Blackwell platform, and customers with limited tolerance for risk often default to the most established option.

OpenAI talks add another twist

Adding to the strategic complexity are reports that Amazon has been in discussions to invest more than $10bn in OpenAI and to supply it with chips and computing power. If that were to materialise, it would be a notable shift in the competitive landscape, given OpenAI’s current ties with Microsoft and the broader scramble among cloud providers to lock in anchor AI customers.

For Amazon, a deal of that scale would be both a vote of confidence in its infrastructure and an acknowledgement that partnerships may matter as much as in-house development in this phase of the AI build-out.

A familiar message from Jassy

Andy Jassy, Amazon’s chief executive and a former head of AWS, has repeatedly argued the company needs to operate with the speed and focus of “the world’s largest start-up”. He has also warned that AI will reduce corporate headcount over time as processes become more automated and routine work is absorbed by software.

That framing is consistent with the internal language used by Galetti, which emphasises flatter structures and fewer management layers. It is also a continuation of a post-pandemic correction. Amazon expanded rapidly during Covid to meet surging demand across online shopping and digital services, then spent the following years pruning roles as growth normalised.

The scale of the latest announcement shows that process is not complete. Amazon eliminated about 27,000 roles in two rounds in 2023, while AWS cut hundreds of jobs in 2024. The new reduction focuses on corporate positions rather than warehouse staff, reinforcing the sense that the biggest changes are in the layers of management and support functions built up during the boom.

What happens next

Galetti stressed that Amazon would continue hiring in “strategic areas” critical to its future, even as it removes thousands of jobs elsewhere. The message, in effect, is that the company wants fewer people doing co-ordination and internal administration, and more focused teams building products and infrastructure that keep Amazon competitive in AI.

For staff, the practical impact will be immediate and personal, particularly for those facing the 90-day internal job search. For investors, the bigger question is whether the drive to cut layers and accelerate decision-making translates into clearer AI wins for AWS, and whether Amazon can prove it has a differentiated offering as the industry’s next growth cycle gathers pace.