BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//pretalx//pretalx.com//bsidesatl-2025//speaker//7LVXGU
BEGIN:VTIMEZONE
TZID:America/New_York
BEGIN:STANDARD
DTSTART:20241115T000000
TZNAME:EST
TZOFFSETFROM:-0500
TZOFFSETTO:-0500
END:STANDARD
BEGIN:DAYLIGHT
DTSTART:20250309T030000
RDATE:20260308T030000
TZNAME:EDT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
END:DAYLIGHT
BEGIN:STANDARD
DTSTART:20251102T020000
RDATE:20261101T020000
TZNAME:EST
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
SUMMARY:From Vibes to Vulnerabilities: How AI Agents Turn on Their Users -
  Wes Widner
DTSTART;TZID=America/New_York:20251115T130000
DTEND;TZID=America/New_York:20251115T135000
DTSTAMP:20260812T151225Z
UID:pretalx-bsidesatl-2025-QXM7UM@pretalx.com
DESCRIPTION:We were promised autocomplete on steroids. What we got was a n
 ew attack surface\, one that developers invite into their terminals\, edi
 tors\, CI pipelines\, and even production systems.\n\nIn this talk\, I wal
 k through how AI coding agents\, the ones we rely on to ship faster by off
 loading mental load\, are quietly introducing a new class of threats. And 
 these aren’t theoretical. They’re already being exploited in the wild.
 \n\nWe’ll explore how natural gaps in agent understanding can become opp
 ortunities for adversaries\, and how the tools built to boost productivity
  can be subverted into delivery mechanisms for exploitation.\n\nFrom subtl
 e context manipulation to unexpected supply chain consequences\, we’ll t
 race how trust in your agent can become the thing that gets you pwned.\n\n
 This isn’t about prompt injection. It's about something much deeper. Thi
 s is real-world exploitation. Where the agent becomes the source of the n
 ext attack.\n\nWe’ll walk through concrete examples\, highlight the (sur
 prisingly limited) tooling available today\, and make the case that agent 
 context and model provenance need to be treated with the same rigor we alr
 eady apply to our dependencies and infrastructure.\n\nAI agents are immens
 ely useful. But if we don’t rethink how we trust and monitor them\, they
  won’t just make our jobs easier\, they’ll make attackers’ jobs easi
 er too.
LOCATION:Room 300
URL:https://pretalx.com/bsidesatl-2025/talk/QXM7UM/
END:VEVENT
END:VCALENDAR
