Today, Amazon joined
President Biden and leaders across government and industry to commit to
continue promoting the safe, secure, and transparent development of artificial
intelligence (AI) technology that benefits society.
The White House met
with leading AI companies to identify commitments related to the safe and
responsible development of future generative AI models that are overall more
powerful than any currently released foundation models (FMs). These commitments
build on Amazon’s approach to responsible and secure AI development, and they
will help pave the way for a future that enhances the benefits of AI and
minimizes its risks. As part of this engagement, Amazon welcomes the
following voluntary commitments to:
1. Commit to internal
and external adversarial-style testing (also known as "red-teaming")
of models or systems in areas including misuse, societal risks, and national
security concerns, such as bio, cyber, and other safety areas.
2. Work toward
information sharing among companies and governments regarding trust and safety
risks, dangerous or emergent capabilities, and attempts to circumvent
safeguards.
3. Develop and deploy
mechanisms that enable users to determine if audio or visual content is
AI-generated, including robust provenance, watermarking, or both, for
AI-generated audio or visual content.
4. Invest in
cybersecurity and insider threat safeguards to protect proprietary and
unreleased model weights.
5. Incentivize
third-party discovery and reporting of issues and vulnerabilities.
6. Publicly report model
or system capabilities, limitations, and domains of appropriate and
inappropriate use, including discussion of societal risks, such as effects on
fairness and bias.
7. Prioritize research
on societal risks posed by AI systems, including on avoiding harmful bias and
discrimination, and protecting privacy.
8. Develop and deploy frontier AI systems to help address society’s greatest challenges.
At Amazon, we are
committed to continued collaboration with the White House, policymakers, the
technology industry, researchers, and the AI community to advance the
responsible and secure use of AI. As one of the world’s leading developers and
deployers of AI tools and services, Amazon supports these voluntary commitments
to foster the development of AI that is safe, responsible, and trustworthy. We
are dedicated to driving innovation on behalf of our customers while also
establishing and implementing the necessary safeguards to protect consumers and
customers.
The White House
commitments are forward-looking and are aligned with Amazon’s approach to
responsible and secure AI development. Amazon builds AI with
responsibility in mind at each stage of our comprehensive
development process. Throughout design, development, deployment, and operations
we consider a range of factors including accuracy, fairness, appropriate usage,
toxicity, security, safety, and privacy. For all of our FMs, we continually
work to improve our features and learn from customers as they experiment with
new use cases.
Responsible use of AI
technologies is key to fostering continued innovation, and Amazon is committed
to developing fair and accurate AI services. For example, Amazon CodeWhisperer is the only AI
coding companion with built-in security scanning for finding and suggesting
remediations for hard-to-detect vulnerabilities. On intellectual property, it
uniquely provides source code provenance analysis and indicates which open source
licensing models may be applicable to generated code. Amazon’s Titan family of foundation
models is also built to detect and remove harmful content in the data that
customers provide for customization, reject inappropriate content in the user
input, and filter the model’s outputs containing inappropriate content such as
hate speech, profanity, and violence. And Amazon SageMaker Clarify detects
and measures potential bias using a variety of metrics so developers can
address potential bias and explain model predictions.
We are also committed
to providing customers with tools and resources to develop and use AI
responsibly. For example, at our re:Invent 2022 conference, we announced the
launch of AWS AI Service Cards, a
transparency resource to help customers better understand our AWS AI services.
The AI Service Cards deliver a form of responsible AI documentation that
provide customers with a single place to find information on the intended use
cases and limitations, responsible AI design choices, and deployment and
performance optimization best practices for our AI services.
The voluntary
commitments also emphasize the important role security plays in responsible AI.
Security is a top priority for Amazon and is foundational in how we design our
AI products and services. All of our services go through rigorous security
testing, and we maintain strict physical, electronic, and procedural safeguards
to protect our systems. We work closely with the security community to further
harden our products and systems, including making it easy to report vulnerabilities.
As one of the most
transformational technologies of our generation, AI provides opportunities to
be a force for good. We look forward to creating innovations for and with our
customers to realize the benefits of AI, as well as continuing our partnership with
the White House to deliver on our shared commitments.
Written by Amazon Sta
