Bitdefender webinar casts unowned AI agents as an insider risk
Original title: The AI Agent Nobody Owns: Watch Expert Webinar On Your Newest Insider Threat
The session cites endpoint research, including a malicious agent skill found in the wild, and argues for governed AI instead of bans.
The Hacker News is promoting a Bitdefender webinar, “AI as an Insider Threat: From Shadow AI to Governed AI,” with Richard De La Torre. It says AI agents, coding assistants, MCP servers, and browser extensions can hold API keys and database passwords without an owner or security review. Bitdefender’s 2026 assessment is cited as finding full AI-tool visibility at 51.8% of organizations, and IBM as finding shadow AI in 20% of breached organizations. Bitdefender Labs research on real machines is also promoted, including a malicious agent skill caught in the wild.
Unconfirmed: The malicious agent skill is mentioned only as a webinar finding, with no technical detail or independent confirmation.
- Only 51.8% of organizations report full visibility into staff AI tools, Bitdefender says.
- IBM found shadow AI involved in 20% of breached organizations, the promo says.
- Bitdefender Labs says it caught a malicious agent skill in the wild on real machines.
- Guest is Richard De La Torre, technical product marketing manager at Bitdefender.
- Vendors
- Bitdefender
- Organizations
- IBMBitdefender Labs
From the original article
2 min read at thehackernews.com
Right now, there is probably an AI agent running on a laptop somewhere in your network. It stores API keys and database passwords, can access your internal systems, and works without anyone watching. Nobody reviewed its access.
thehackernews.com
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.