TLDR IT 2026-05-01

Google Cloud crossed $20B in quarterly revenue, up 63%, as enterprise AI demand continued to push cloud spending higher. Alphabet also spent $35.7B in capex, showing how aggressively hyperscalers are racing to add AI capacity.

After an AI agent accidentally deleted a production database using a long-lived token, Railway introduced a 48-hour soft delete policy for all API volume deletions. The platform is now prioritizing granular token permissions and agent-specific interfaces like MCP to ensure destructive actions remain reversible and secure.

Microsoft Security has launched a new update series focused on helping security and IT teams secure AI agents, strengthen core infrastructure, and respond to threats in real time. The updates reflect how agent governance is becoming a core enterprise security concern, not just an AI experimentation problem.

Australia's financial regulator warned banks that frontier AI systems could help attackers find and exploit vulnerabilities faster. The regulator also flagged board-level AI literacy and vendor overreliance as weak spots in financial-sector risk controls.

Autonomous business workflows are forcing companies to rethink how work gets done, not just where AI can be inserted. Leaders need to focus on the specialized human value employees bring as agents take over more operational tasks.

Claude Security is now in public beta for Claude Enterprise customers. The tool finds code vulnerabilities and generates fixes. The release comes as AI-powered exploit discovery raises pressure on defenders to adopt more automated security tooling.

Wiz has launched new capabilities to secure AI-native development, including an AI-BOM for framework inventory and IDE plugins for real-time guardrails. These tools map code-to-cloud risks, enforce OWASP-aligned SAST rules, and enable automated remediation, helping security teams maintain visibility and control as AI agents accelerate software delivery.

Salesforce is crowdsourcing parts of its AI roadmap with customers as enterprise software vendors face pressure to ship faster. The approach reflects a broader shift toward tighter customer feedback loops for enterprise AI products.

Arc is an internal agentic AI platform that lets Citi's employees securely build and deploy AI agents across the company. The platform gives teams access to multiple leading AI models in one system, with agents able to handle tasks like portfolio data compilation, market trend analysis, and scenario testing.

Microsoft's April 2026 KB5083769 security update is causing failures in third-party backup applications on Windows 11 24H2 and 25H2 systems.

Sandhills Medical Foundation disclosed a ransomware-related breach that affects nearly 170,000 individuals.