C1

If our IGA Haunted House didn't scare you, these answers from the team might. 👻 Happy Halloween from C1.ai!

Tired of building roles from entitlement spreadsheets and tribal knowledge? With one click, Role Mining Suggestions show admins which groups are worth turning into access profiles. They also keep roles current as teams change. See how it works → https://okt.to/dBHvpw

This Halloween, your outdated identity program is handing out tricks like privilege creep, orphaned accounts, and admin access that never expires. 👻 The treat is C1.ai, which cleans out every haunted room so you can stop checking under the bed. Swipe for the full tour! 🎃 🍬

Teams that required formal external approval were 2.6 times more likely to be low performers, according to DORA's 2019 research, and the process didn't reduce change failures. That finding shapes a big part of The new paved road: https://okt.to/URTxWj As more non-engineers ship apps, the paper argues against rebuilding the old gates. Controls that prevent a specific harm stay, process that produces no useful outcome goes, and the evidence auditors need gets generated automatically while the road runs.

Some ghosts are harmless. The ones that still have access to your production systems are a different story. Happy Halloween from the team at C1.ai, and may all your old accounts rest in peace 👻 https://okt.to/6URMzh

This is what the first ever C1 Transform looked like from inside the room. One day in San Francisco with the security leaders of SpaceX, Anthropic, NVIDIA, Vercel, Brex, Vanta, and more, talking candidly about what it takes to govern AI agents in the enterprise. Consider this your preview, because Transform is coming back in 2027 and the room is only getting bigger. #C1Transform #AgenticAI #identitysecurity

“Pure play software is kind of deterministic, mostly trustable. Humans, non-deterministic, trustable. Agents, non-deterministic, non-trustable.” Our CEO Alex Bovee laid out this framework on The Non-Human & AI Identity Podcast, along with what replaces trust when an agent can’t earn it: scope each agent’s mission, then authorize every action against that scope at runtime. Read the full breakdown: https://lnkd.in/gxrXiEYF #AIAgents #IdentitySecurity NHI Mgmt Group

Is an AI agent an employee or a piece of software? Our CEO Alex Bovee says it is both, and that is exactly why agents are so hard to secure. This edition shares how he thinks about giving them access without having to trust them.

Every team deploying AI agents hits the same wall: how much autonomy should an agent have? Our co-founder and CEO Alex Bovee is digging into exactly that on the “Giving agents more autonomy without losing control” panel at the Agentic AI Summit in Boston, October 29 at 12 PM ET, alongside AI leaders from The Mutual Group and LG Electronics. Catch the session, then find the C1 team at booth 10. https://okt.to/fbGlMH #AgenticAI #AISecurity #AIAI AI Accelerator Institute

What happens to your security review when nobody buys the software? For years, privacy review, vendor assessment, and procurement kicked in the moment someone wanted to purchase a tool. Now the team that needs it can build it in an afternoon, and none of those reviews ever start. The new paved road sorts what went missing into dropped, orphaned, and new responsibilities, and lays out a different fix for each: https://okt.to/FB5vKT