Security | Threat Detection | Cyberattacks | DevSecOps | Compliance

Security teams are dealing with a fundamentally different operating environment than they were a few years ago. AI-assisted development is rapidly pushing more code and infrastructure into production, and according to Datadog’s 2026 State of DevSecOps report, 40% of running services have an exploitable vulnerability.

If you’re like our team, the morning after the Claude Mythos announcement brought more questions than answers. Among them: “Serious question. Do customers still need SAST?” It’s a fair question if you stop at the headline. Claude Mythos, Anthropic’s frontier AI model currently gated to vetted partners through Project Glasswing, had autonomously identified thousands of zero-day vulnerabilities across major operating systems and browsers . No rule books, no checklists.

By mid-2026, the question is no longer whether AI belongs in the enterprise. It’s already embedded in daily work, supporting research, development, customer engagement, and operations. AI agents now act on behalf of employees, automate decisions, and interact directly with enterprise data and systems. This shift creates a new security challenge.

On April 27, 2026, a threshold was crossed that the internet had never hit before. Cloudflare Radar data confirmed that automated systems, such as bots, crawlers, and autonomous AI agents, now generate 57.4% of all HTTP requests for web content. Human traffic accounts for just 42.6%. What is accelerating this transformation is agentic AI: autonomous systems that browse, search, authenticate, and transact on behalf of users without any human intervention mid-task.

Employees are adopting AI tools, agents and automations faster than organizations can govern them. The real danger emerges when these tools connect directly to internal systems and sensitive data in the name of enhancing productivity. Among employees who use AI at work, a significant share do so without formal approval from IT or security teams, which is commonly called shadow AI.

What recent AI security developments should change and what they should not—for security leaders and vulnerability management teams The question is no longer whether AI will affect vulnerability discovery. It is whether exposure management programs can keep up with a faster clock.

ThreatSpike exhibited at Infosecurity Europe 2026 at ExCeL London from 2–4 June this year. Three days, ten expert sessions presented on our stand and more conversations about AI than we’ve had at any event in recent memory. This is our round-up: what we saw on the floor, what we presented, and what the industry is clearly wrestling with right now.

AI demos are easy. AI you’d actually trust near your control environment is not. If you’ve sat through a few of these pitches lately, you’ve probably landed on the same four questions every CISO we talk to is asking. And you’re right to ask them.