As the AI era officially unfolds, profound changes in social production and human lifestyles have emerged, alongside escalating security concerns. Balancing technological innovation with security and trustworthiness while exploring the deep application of generative AI has become a crucial topic in driving the steady development of the AI industry.
On June 12th, at the re:Inforce 2024 global conference, Amazon Cloud Technology announced a series of new features for security services, covering aspects such as malware protection, AI-driven security, and identity access and management.
Chris Betz, Chief Information Security Officer of Amazon Cloud Technology, believes that generative AI can provide powerful tools to everyone, profoundly altering almost all types of customer experiences. However, without clear governance, generative AI can lead to security and privacy issues. “Employees enthusiastic about generative AI often see security as an obstacle. This is not only a mistaken mindset but also harmful to the business. Amazon Cloud Technology has always believed that security is a business driver. Security reduces risks, enhances resilience, and empowers customers to innovate more quickly and confidently.”
Specifically, Amazon Cloud Technology has further extended the malicious protection features of Amazon GuardDuty to Amazon Simple Storage Service (Amazon S3) to provide comprehensive security protection. They also introduced the preview version of the natural language query feature based on generative AI for Amazon CloudTrail Lake, enabling users to query and analyze cloud resource usage more conveniently.
In practical terms, Amazon Cloud Technology uses Sonaris to analyze network traffic, identify, and block attempts by a large number of customer accounts seeking vulnerabilities. Chris Betz mentioned that from May 2023 to April 2024, Sonaris denied over 24 billion attempts to scan data stored in Amazon Simple Storage Service (Amazon S3) by customers and blocked nearly 2.6 trillion attempts to exploit services running on customer Amazon EC2 virtual servers.
In terms of identity access and management, Amazon Identity and Access Management (Amazon IAM) added support for using passkeys as a second factor for authentication, enhancing account security. Additionally, Amazon IAM Access Analyzer provides recommendations for unused access permissions to help users implement the principle of least privilege and provides policy access support for public and critical resources.
Furthermore, Amazon Cloud Technology also introduced service embedding features for Amazon Cloud WAN and support for the Simple Certificate Enrollment Protocol (SCEP) for Amazon Private Certificate Authority (Amazon Private CA), further enhancing the security and convenience of its cloud services.
“Security should be built into every product roadmap, engineering plan, and weekly meeting, just like features, performance, and costs. The best security is not added at the end of the process or ‘bolted on’ externally to the system; instead, security should be an indispensable foundation,” said Chris Betz.
“We must ensure the security of users’ data directly and also the security of the data of those enterprises’ end-users who use services provided by Amazon Cloud Technology,” Chris Betz stated. “Cybersecurity challenges will continue to evolve, and Amazon Cloud Technology is committed to progressing by innovating and improving its technology and security culture.”
