Amazon expands internal AI tools to 700+ teams as adoption accelerates

Amazon is expanding the use of its internal AI tools across more than 700 teams, as the company intensifies efforts to embed AI into everyday engineering workflows. The move reflects a broader push to drive productivity gains at scale, even as parts of its workforce express concerns about how the rollout is being managed.

According to an internal document reviewed by Business Insider, Amazon’s retail division is systematically tracking how engineers adopt AI, how frequently tools are used, and whether this translates into measurable output. The initiative places AI at the centre of how software is built, tested, and deployed across the organisation.

Scaling AI across engineering workflows

Amazon’s AI push is both wide-ranging and tightly managed. The company expects a significant majority of its engineering teams to adopt what it describes as “AI-native” ways of working.

Key indicators from the internal document include:

More than 700 teams are actively using AI tools such as AI Teammate
Around 60% of retail engineering teams had adopted AI practices as of February
The company expects 80% adoption across these teams over time
Over 2,100 engineering teams are being guided to integrate AI into workflows

AI Teammate, one of Amazon’s key tools, integrates with workplace systems to automate tasks by analysing chats, documents, and tickets. Other tools, including Pippin, which converts ideas into technical designs, and coding assistant Kiro, are also seeing growing use, the document noted.

A spokesperson for Amazon told Business Insider that integrating AI across the full development lifecycle, rather than using it selectively, has delivered the most meaningful gains in speed and innovation.

Productivity goals shape adoption strategy

Amazon is linking AI adoption directly to engineering productivity targets. The internal document outlines ambitious expectations for teams:

Most teams are expected to triple software release velocity
A smaller group of at least 25 teams aims for tenfold output gains
Progress is tracked by the company’s senior leadership group, known as the S-Team

The document advises managers to treat AI like any automation investment, encouraging teams to actively identify use cases, measure outcomes, and build repeatable practices.

At the same time, Amazon is monitoring performance through detailed metrics, including tool usage rates, monthly active users, and output-linked indicators such as “Value Deriving Events”, which measure actions like generating outputs or providing feedback.

Measurement focus and evolving approach

Amazon’s approach places strong emphasis on measurement, but with caution. The company explicitly references “Goodhart’s Law”, the idea that when a metric becomes a target, it can distort behaviour.

To address this, the company tracks both access to AI tools and actual usage, alongside employee sentiment indicators such as Net Promoter Scores, according to Business Insider.

“Set clear adoption and engagement targets,” the document states, while also urging teams to balance measurement with meaningful outcomes.

Internal friction and course correction

Despite rising adoption, the rollout has generated internal resistance. Feedback cited in the document points to concerns around centrally driven mandates, unclear success metrics, and the burden of self-reporting progress.

Engineers have also highlighted practical challenges, including complex onboarding processes and duplication of tools, sometimes referred to internally as AI sprawl.

In response, Amazon is adjusting its approach:

Moving towards collaborative AI practices rather than strict mandates
Replacing manual reporting with automated tracking systems
Developing a centralised learning platform to share best practices
Allowing teams more flexibility in choosing tools and implementation methods

An Amazon spokesperson said the company does not centrally mandate specific AI tools and encourages teams to adopt what works best, adding that internal debate is part of its operating culture.

A shift towards AI-native work

Amazon’s internal framework for AI adoption prioritises speed, flexibility, and practical outcomes. The company emphasises using AI where it adds value, rather than forcing it into every process, while ensuring systems remain understandable and auditable.

Engineers are encouraged to integrate AI into daily workflows and identify tasks where automation can improve efficiency. Managers, in turn, are expected to provide clarity, access, and direction.

Amazon’s expanding AI deployment highlights a broader shift in how large organisations approach technology adoption. The focus is moving beyond experimentation to embedding AI into core workflows and performance metrics.

However, the company’s experience also underscores a parallel reality. Adoption at scale requires not just tools, but alignment on measurement, usability, and employee buy-in. As enterprises push for AI-led productivity gains, balancing these elements is likely to remain a defining challenge.

AI & Emerging Tech