Breaking

IBM Exec Urges Fighting AI Attacks With AI as Public-App Exploits Jump 44%

June 25, 2026 at 09:54 EDT

Security
AI Agents
Sales & Business

Gaurav Agarwal, Vice President, Technology at IBM India & South Asia, said organisations must counter increasingly automated and faster AI-driven cyberattacks with AI-powered defences, citing the limits of traditional perimeter security and the surge in attacks documented in the IBM X-Force Threat Intelligence Index 2026.

IBM Industry Perspective · Cybersecurity

"Fight AI with AI": Attackers Are Scaling, Defenders Must Too

Cybercriminals now use AI to automate vulnerability discovery, exploit creation, and attacks at speed — outpacing human-only defenses. IBM data shows organizations that deploy AI defenses pay far less when breaches strike.

The Defender's Dividend — average breach cost

Same threat, two outcomes. Extensive AI & automation use vs. none.

$3.62M

Extensive AI & automation

$5.52M

No AI / automation

~$1.9M gap · breaches contained ~80 days faster with AI defenses

$4.44M

Global avg. breach cost (down 9% YoY)

16%

of breaches involved attacker AI

~$670K

extra cost from Shadow AI

~⅓

of orgs use AI defenses extensively — adoption has stalled

Why the old wall fails — identity is the new perimeter

Cloud, remote work & AI agents

→

Perimeter walls lose value

→

Govern identity & least privilege

Top intrusion paths: public-facing apps & valid credentials. New agent risks: rogue prompts and Shadow AI — calling for AI governance and human-in-the-loop oversight on critical decisions.

Project Lightwell · supply-chain security (May 28, 2026)

$5B

IBM + Red Hat commitment

20,000+

engineers paired with AI

~3,900

high/critical open-source flaws found by Mythos Preview

Continue reading

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