Microsoft: AI Makes Cyberattacks Cheap, Demands Safer Code

Microsoft's David Weston warned at Black Hat USA 2026 that AI has ended the era of scarce vulnerabilities, making powerful offensive tools widely accessible. Traditional defenses built on the assumption that attacks are rare and costly no longer hold, he argued. Weston called on security teams to reinvest AI-driven productivity gains into safer software construction, including memory-safe languages like Rust and formal verification methods. Google's shift to safer languages cut memory-safety vulnerabilities from 75% of its total in 2019 to under 20% by 2025, with zero new memory-safety issues across 5 million new lines of code.