Leaked Amazon documents reveal 9 top factors customers consider before buying AI models and services

• Amazon's internal guidelines help AWS salespeople address questions about AI rivals such as OpenAI.
• The guidelines highlight nine factors customers consider when buying generative-AI models.
• AWS wants to focus on cloud infrastructure and AI foundation models, not AI chatbots.

Amazon's internal sales guidelines give talking points for dispelling hype around OpenAI and answering customer questions about Microsoft's and Google's competing AI products.

The guidelines, which Business Insider obtained, also list nine important factors customers consider before purchasing generative-AI models and services.

The instructions, mainly used by AWS salespeople, provide clues about the company's AI priorities. The nine criteria include security, cost, and the ability to personalize models using a technique called retrieval-augmented generation, or RAG.

Here are the nine factors:

• Customization: The ability to tailor AI models for specific requirements (e.g., the style of a model's outputs).
• Personalization: The ability to use in-house company data to make AI-model outputs more relevant and tailored (i.e., fine-tuning or RAG).
• Accuracy: How closely the generated output aligns with desired objectives.
• Security: The implementation of appropriate measures for data and privacy protection.
• Monitoring: The ability to identify issues like drift, bias, or degradation in output quality.
• Cost: The overall expenses, including initial investment and ongoing costs associated with training, deployment, maintenance, and infrastructure.
• Ease of use: The model's usability, integration capabilities, and availability of support.
• Responsible AI: The model's adherence to ethical guidelines and ability to address biases, provide explanations for outputs, and incorporate safeguards against misuse.
• Innovation: A service provider's perceived status as an innovator relative to others.

The guidelines tell AWS salespeople to drive their pitches toward the foundation models and cloud infrastructure needed to build AI services instead of focusing too much on the popularity of AI chatbots. This is because AWS's core strength is in cloud infrastructure, not in consumer AI chatbots, though the company is working on its own ChatGPT competitor.

One of the guideline documents calls out AWS's "Value Propositions" that should be tapped when selling to prospective customers. These include AWS's ease of use for building and customizing AI offerings and its robust security and privacy features. The document also mentions AWS's "price-performant infrastructure," including its own AI chips and AI applications built by AWS, such as Amazon Q.

"It's no secret that generative AI is an extremely competitive space. However, AWS is the leader in cloud and customer adoption of our AI innovation is fueling much of our continued growth," an AWS spokesperson told BI in an email.

"AWS has more generative AI services than any other cloud provider, which is why our AI services alone have a multi-billion dollar run rate," the spokesperson added. "It's still early days for generative AI, and with so many companies offering varied services, we work to equip our sales teammates with the information they need to help customers understand why AWS is the best, easiest, most performant place to build generative AI applications. To parse the language as anything more than that or mischaracterize our leadership position is misguided speculation."

Do you work at Amazon? Got a tip?

Contact the reporter, Eugene Kim, via the encrypted-messaging apps Signal or Telegram (+1-650-942-3061) or email (ekim@businessinsider.com). Reach out using a nonwork device. Check out Business Insider's source guide for other tips on sharing information securely.