Breaking

Index Ventures Partner Warns AI Attacks Could Push Cybersecurity Toward Machine-to-Machine Defense

September 23, 2026 at 16:34 EDT

Security
AI Agents
Sales & Business

Index Ventures partner Shardul Shah is arguing that cybersecurity is entering a machine-versus-machine phase, where periodic human review and manual response may no longer keep pace with AI-enabled attackers. In a new Equity podcast episode, Shah said the security stack could be remade as AI lowers the cost and speed of offensive activity, including a scenario in which one attacker deploys a 10,000-agent ransomware swarm.

Sept. 23, 2026 · AI security

Cyber defense faces machine-speed attacks

Index Ventures partner Shardul Shah says AI-enabled attacks could force security teams beyond periodic human review and toward machine-to-machine defense.

10,000

agents in a ransomware-swarm scenario

$32B

Google acquisition of Wiz

650K+

7AI security investigations reported

A 10,000-agent swarm, shown as 10 blocks

10,000 AI agents

Each block represents 1,000 agents

Why manual queues struggle

AI-assisted reconnaissance

→

Automated exploit research

→

Agentic investigation and response

Defensive promise

Security teams see agentic tools as a way to triage alert overload, investigate where data lives and reserve people for strategic threat work.

Deployment risk

Guidance urges low-risk tasks first, tight permissions and no broad access to sensitive systems or critical data.

The emerging stack

Autonomous triage
Agents investigate repetitive alerts.

Human judgment
People handle risk, architecture and escalation.

Guardrails
Least privilege and continuous oversight define the boundary.

Continue reading

The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.