Safe Code Warrior has signed a strategic collaboration settlement with Amazon Net Companies and launched Amazon Bedrock coaching modules on its platform.
The brand new materials is geared toward builders and engineers working with generative AI functions.
The settlement connects Safe Code Warrior’s coaching enterprise with AWS’s push round Amazon Bedrock, its service for constructing generative AI functions and brokers. The brand new modules are designed to assist growth groups undertake safe practices when utilizing Bedrock.
Safety focus
The coaching focuses on securing infrastructure as code for Amazon Bedrock by way of Terraform. It’s supposed to reveal customers to frequent safety points in AI and enormous language mannequin deployments, together with immediate injection, extreme company, inadequate logging, and knowledge publicity.
The package deal contains 4 coding labs, 4 AI challenges, and one walkthrough mission. Delivered by way of the Safe Code Warrior platform, the workouts give builders direct apply in figuring out and addressing safety weaknesses in managed environments.
The launch displays rising demand for sensible safety coaching as corporations introduce generative AI instruments into software program growth and customer-facing methods. Many have moved from restricted testing to wider deployments, elevating considerations about how builders handle dangers tied to mannequin behaviour, knowledge dealing with, and system entry.
Pieter Danhieux, Chief Government Officer and Co-founder of Safe Code Warrior, mentioned the corporate sees a spot between curiosity in AI deployment and the extent of hands-on safety information inside many growth groups.
“Software program growth is being augmented by generative and agentic AI expertise on an unprecedented scale, and this strategic collaboration couldn’t come at a extra mission-critical second for each safety leaders and future-focused builders,” mentioned Danhieux.
“As organizations proceed to undertake generative AI functions and instruments, an inherent degree of danger is launched, however this danger dramatically will increase when builders should not correctly educated on the instruments and platforms they’re leveraging. Builders want greater than theoretical steerage; they require sensible expertise in figuring out and mitigating real-world threats. By getting into into this SCA with AWS to ship hands-on coaching for Amazon Bedrock, we’re serving to enterprise groups study to confidently tackle AI vulnerabilities and crimson flags in order that they will construct AI functions which can be safe by default,” added Danhieux.
Buyer view
Safe Code Warrior additionally cited OpenText as an early consumer of Amazon Bedrock in its growth and product work. The shopper reference highlights how safety coaching distributors are attempting to tie their merchandise extra intently to the cloud companies and AI methods corporations already use.
“Amazon Bedrock is a foundational service for each our AI-assisted growth and AI-powered product performance throughout platforms. Safe Code Warrior’s new interactive Bedrock coaching modules empower our builders with the hands-on expertise to additional strengthen our potential to forestall AI dangers like immediate injection and knowledge leakage. Proactive controls allow our groups to study to construct secure-by-design habits from the beginning, accelerating our AI roll out right into a production-ready setting with out sacrificing safety or software program high quality,” mentioned Dylan Thomas, Senior Director of Engineering at OpenText.
AI danger coaching
Safe Code Warrior, based in Australia and working throughout a number of markets, has constructed its enterprise round safe coding schooling for software program builders. Extra just lately, it has expanded its focus to governance and oversight of AI-assisted software program growth, an space that has grown as coding assistants and AI-driven growth instruments have turn out to be extra frequent in engineering groups.
The Bedrock modules are one other instance of safety corporations creating AI-specific coaching slightly than adapting older software safety programs. Dangers resembling immediate injection and knowledge leakage differ from conventional coding flaws as a result of they usually emerge from mannequin interactions, system prompts, permissions, and the way in which functions join fashions to exterior instruments and knowledge sources.
That shift has created a marketplace for coaching that mixes software program safety fundamentals with operational steerage for AI companies. It additionally displays stress on corporations to indicate that builders perceive the dangers of deploying AI methods in manufacturing, significantly in regulated sectors and enormous enterprises the place governance necessities are rising.
Platform technique
For AWS, the collaboration provides Bedrock customers entry to coaching constructed round sensible situations in a platform already utilized by developer groups. For Safe Code Warrior, it offers a method to place itself extra immediately within the rising marketplace for AI safety schooling tied to particular cloud merchandise.
The brand new content material centres on developer habits and decision-making slightly than automated safeguards alone. That emphasis suggests safety coaching suppliers see human judgement as a unbroken weak level in AI roll-outs, at the same time as cloud distributors add extra controls and coverage instruments round mannequin use and deployment.
The modules can be found now by way of the Safe Code Warrior platform and are designed to simulate remediation of real-world AI dangers for Amazon Bedrock customers.