Cloud trust is becoming the next cybersecurity battleground for AI
Cloud trust is becoming the next cybersecurity battleground for AI
Publish Date: 2026-07-03 16:58:00
Source Domain: www.digitaljournal.com
As artificial intelligence (AI) workloads continue their rapid migration to the cloud, a new question is emerging for enterprises, regulators, and cybersecurity professionals: how can organisations verify that the infrastructure processing sensitive data has not been altered, compromised, or tampered with?
The issue is attracting increasing attention as businesses shift mission-critical workloads including AI models, financial systems, healthcare applications, and sensitive analytics to public cloud environments. While cloud providers invest heavily in security, many organisations are seeking independent ways to verify the integrity of the underlying infrastructure.
A new initiative from cybersecurity company DigiCert, developed in collaboration with Google Cloud, aims to address that challenge through the use of independent cryptographic attestation for confidential computing environments.
For years, organisations have relied on cloud providers’ security controls and assurances. However, as data privacy regulations tighten and AI adoption accelerates, customers increasingly want evidence rather than assurances.
The challenge becomes even more complex with generative AI. Large language models process vast amounts of proprietary data, and organisations need confidence that information is being handled within secure and verified environments. According to DigiCert, the solution lies in extending the principles of Public Key Infrastructure (PKI), the technology that underpins trusted internet communications, to cloud infrastructure itself.
What is confidential computing and do you need it?
Confidential computing is one of the fastest-growing areas in cybersecurity.
Unlike traditional cloud security, which protects data when it is stored or transmitted, confidential computing protects information while it is actively being processed. This is achieved through hardware-based trusted execution environments that isolate workloads from other…
