AI agent swarm breaches 100+ sites: lessons for Indian firms

The News

A single attacker used a swarm of largely autonomous AI agents to compromise more than 100 websites, according to research published by AI security firm Norforge. The firm did not name the victims, but Reuters has reported that they included John Deere, SharkNinja and Cummins, with targets spread across the US, Canada, Australia and Japan.

Norforge found the campaign by examining exposed deployments of Cairn, an open-source agent platform that bills itself as a "general-purpose problem-solving engine". Users give Cairn's agents a starting point and a goal, and the agents log their progress as they work. The researchers identified more than 20 exposed Cairn instances they consider likely malicious. The most active one belonged to this attacker, who set out preferred intrusion tactics in fewer than 100 Chinese characters. Successful breaches took roughly 23 hours each, per Reuters.

The companies have played down the impact. John Deere said there was "no indication that sensitive information was accessed". Cummins said it was investigating a possible attack on a site run by one of its third-party vendors, and SharkNinja said a standalone product-support site was affected and the threat had been contained. Cairn's creator, known online as @le1xia0, has argued that misuse "is a human decision, not a feature of the engine".

Why It Matters

The striking detail is not the victim list but the economics. A brief instruction set and an open-source orchestration layer were enough to run a multi-country campaign with minimal human supervision. That collapses the cost of reconnaissance and exploitation, which were historically the labour-intensive parts of an intrusion.

There is a recent precedent. In November 2025, Anthropic disclosed that it had disrupted a campaign it attributed to a Chinese state-sponsored group that used its Claude Code tool to target around 30 organisations, with the AI performing most of the operational work. The Norforge findings suggest the same pattern is now available to individual operators using freely downloadable software, not just well-resourced state teams.

The Cummins case also underlines a familiar weak point: third-party vendor sites. Automated agents can probe peripheral assets that security teams rarely prioritise.

Indian Angle

Two of the named companies have deep Indian footprints. Cummins India Ltd is listed on the NSE and BSE, and John Deere runs tractor manufacturing and a large technology centre in Pune. Nothing reported so far indicates that Indian operations were affected, but the episode is a reminder that a vendor-hosted microsite anywhere in a group's estate can become a reputational and disclosure problem for its listed subsidiaries.

For Indian firms, the regulatory stakes are concrete. CERT-In's April 2022 directions require cyber incidents to be reported within six hours of being noticed. The Digital Personal Data Protection Act, 2023 allows penalties of up to Rs 250 crore for failing to take reasonable security safeguards, and its implementing rules are now being phased in. Regulated entities face further layers: RBI's Master Direction on IT governance and SEBI's Cybersecurity and Cyber Resilience Framework both place explicit obligations on managing third-party and vendor risk.

Indian boards and CISOs should treat agent-driven attacks as a volume problem. If one operator can probe hundreds of sites in a day, asset inventories that miss vendor-run subdomains, support portals and campaign pages are the gap that matters most.

FAQ

Was customer data stolen?

The companies say the impact was limited. John Deere said there was no indication sensitive information was accessed, SharkNinja said only a standalone product-support site was hit and contained, and Cummins said it was investigating a reported attack on a vendor-operated site. Investigations may still change that picture.

Are Indian companies required to report incidents like this?

Yes. Under CERT-In's 2022 directions, service providers, intermediaries, data centres and body corporates in India must report specified cyber incidents within six hours of noticing them. Banks, NBFCs and SEBI-regulated entities have additional reporting duties to their regulators, and personal data breaches carry obligations under the DPDP framework.

What should Indian businesses do now?

Start with a full inventory of internet-facing assets, including sites run by agencies and vendors on your domains. Enforce patching and authentication standards in vendor contracts, monitor for automated scanning patterns, and rehearse the six-hour CERT-In reporting process so that legal, IT and communications teams can act quickly.

This story was reported by The Verge. Read the full original coverage at The Verge.