Keycard

Keycard puts users, developers and companies in control of AI agents with real-time, adaptive permissions and complete transparency.

[Keycard @KeycardAI]

Watch Kim's talk: https://t.co/2YoAmzhIzg

[Keycard @KeycardAI]

"It's 10pm. Do You Know Where Your Agents Are?"

That's the title of the talk our Head of DevRel, @KimMaida, gave at @aiDotEngineer World's Fair.

Link below to watch the recording, now LIVE

[AI Engineer @aiDotEngineer]

🆕 Live now: our entire AI x Security Track!

- @Steve_Yegge, Legendary AI Engineer/Speaker
- @mnair1, CTO/CIO Snyk
- @eugeneyan, MTS @AnthropicAI
- @KimMaida, Founding GTEM @keyboardlabs
- @ethansutin, Head of AI, @bee__computer
- Aaron Stanley, CISO @DBT
- Lovina Dmello, NVIDIA
- Moritz Johner, Form3
- Ezra Tanzer, Snyk

thanks also Ethan Cha of @OneCarlyle for being our first Private Equity speaker

full track drop below

[Keycard @KeycardAI]

Watch the full loops debate @aiDotEngineer

https://t.co/cwugWV4Lpi

[Keycard @KeycardAI]

A good cost conscious loop has to track state to know what it's already tried. That state typically exists in a shared memory store.

@ianlivingstone explains "our access control systems weren't designed for this world where machines are acting and reasoning on our behalf"

[Keycard @KeycardAI]

We'll be continuing the conversation on loops, software factories, and securing coding agents at @BlackHatEvents with @Docker and @snyksec on August 3rd

RSVP: https://t.co/wkU5VeItEe

[Insecure Agents Podcast @insecureagents]

The technology that makes software factories possible is here.

The secure design pattern for coding agents is what's missing. Join us at Black Hat for a panel with @Docker @snyksec @KeycardAI to discuss

[Keycard @KeycardAI]

Watch the full debate here: https://t.co/33QsGdd0F6

[Keycard @KeycardAI]

Our CEO @ianlivingstone defended loops during a debate on the main stage of @aiDotEngineer World's Fair

His key points
> software is inherently verifiable
> loops are at the core of software
> at some point a human has to be attributable for an agent's actions

[AI Engineer @aiDotEngineer]

🆕 The Great Loops Debate!

https://t.co/J15hrMVpqi

Team No Delta
- @ianlivingstone
- @GeoffreyHuntley

Team Delta
- @dexhorthy
- @grichadev

led by the inimitable @vtahowe!

Our first ever Oxford Style Debate: There is, or is not, a delta between the hype behind loops and what actually works in practice.

[Keycard @KeycardAI]

In the PocketOS incident a credential meant to manage custom domains was used to delete a database and all its backups.

Diana Kelley CISO @NomaSecurity explores the shared security model for AI and who is responsible.

Our take: that token should never have been able to do it.

[Insecure Agents Podcast @insecureagents]

In the PocketOS incident a database was deleted in 9 seconds.

"Who's responsible for that? Is it the person who gave the API key to the agent? Is it the vendor that trained the AI that came out with the inference to do this destructive task? Is it the platform owner for not putting controls in place?" - Diana Kelley, CISO @NomaSecurity

Diana came on the podcast at @AICouncilConf to discuss the shared security model for AI.

[Keycard @KeycardAI]

Read the full blog post, including how we gate a deploy workflow so credentials only appear after authorization succeeds

https://t.co/BRvpJ7TWaa

[Keycard @KeycardAI]

This week we published a blog post on how we use Keycard to secure our own CI/CD

To set the stage, in March 2025, one compromised GitHub Action leaked secrets from 23,000+ repos.

A year later, the Megalodon campaign hit 5,500 repos in six hours.

Same root cause both times 🧵

[Keycard @KeycardAI]

A job only ever sees the one resource it was issued.

If policy says no, the credential never exists.

[Keycard @KeycardAI]

Our CTO @jaredhanson showed a demo of intent-based authorization (AAuth missions) as a way to reduce consent fatigue and defer credential issuance until after an agent has discovered tools.

[Keycard @KeycardAI]

We were glad to be part of the big tent that came together around agent auth at AAuth Night during @aiDotEngineer World's Fair.

No single layer solves agent auth alone, which is why the whole industry showed up in one room.

[Insecure Agents Podcast @insecureagents]

Recordings from AAuth Night: Moving Beyond OAuth on July 1 during @aiDotEngineer are now LIVE

Watch the presentations, panels, demos and sign up for our next AAuth Night this fall

Also checkout the recap podcast with @DickHardt, Founder of AAuth filmed this morning

[Keycard @KeycardAI]

During the panel our CEO @ianlivingstone shared building great identity and auth infrastructure means making makes use of, and helping define, the best practices available today, while building for what's coming tomorrow.

https://t.co/Dr98UiI7Zd

[Keycard @KeycardAI]

"There's no separation between the control and data planes." @guypod (founder of @tessl_io and @snyksec)

That's why prompt injection works, and why you can't secure a skill from inside the model.

The fix: put the control plane where the context can't reach it.

[Insecure Agents Podcast @insecureagents]

"There's no separation between the control and data planes." @guypod (founder of @tessl_io and @snyksec)

That's why prompt injection works, and why you can't secure a skill from inside the model.

The fix: put the control plane where the context can't reach it.

[Keycard @KeycardAI]

Two CISOs from @box and @NomaSecurity on the security panel at @AICouncilConf described the same problem: teams today can't see half of what's running in their own environments.

You can't govern an agent you can't identify. Identity comes first.