Insecure Agents Podcast

AI engineers and security practitioners listen to us to learn how to give their agents the security they need to reach full autonomy and capability.

[Insecure Agents Podcast @insecureagents]

Full episode available wherever you get your podcasts

https://t.co/pSL27MrwqV

[Insecure Agents Podcast @insecureagents]

"Since you're the founder of OAuth I was wondering, why don't we just extend OAuth to fix agent auth?"

"Well, because it wasn't designed for these things. And so by the time you go and you make all the changes you want to make, it isn't OAuth at all." @DickHardt

[Insecure Agents Podcast @insecureagents]

Dick will be joining us at AAuth Night on July 1st during AIE World's Fair. Come join us to learn more about AAuth and how you can run agents without API keys
https://t.co/cERcO6WOal

[Insecure Agents Podcast @insecureagents]

The future of auth for agents is here

@DickHardt, founder of OAuth, tells us why OAuth isn't working for agents and why we need a new protocol he's working on called AAuth

Full episode available now wherever you get your podcasts

[Insecure Agents Podcast @insecureagents]

Big news!! We've joined @KeycardLabs 🎉

We're excited to continue to grow the Insecure Agents Community and advance the industry together.

[Allie Howe @vtahowe]

https://t.co/oAC8qGT2qc

[Allie Howe @vtahowe]

Big news tomorrow 👀
https://t.co/M0nOZJPQOW

[Insecure Agents Podcast @insecureagents]

For those building with MCP at scale some big changes just dropped

The protocol is now stateless making horizontal scaling, routing, and tool discovery easier.

MCP continues to improve and deliver value as @ianlivingstone predicted during the MCP debate at AIE Code in November

[Ian Livingstone @ianlivingstone]

Session-less MCP is a great step forward - some really nice changes here.

The vibe shift back to MCP is coming, especially as internal app building starts to take off.

[Keycard @KeycardLabs]

Great night at the @insecureagents event at @AICouncilConf. Our co-founder @ianlivingstone was on the panel with @sentry, @browserbase, and @Cloudflare talking about why identity is the bottleneck for agents, followed by a packed happy hour. Thanks to Insecure Agents for putting this together and thanks to everyone that came out.

[Insecure Agents Podcast @insecureagents]

@KeycardLabs @AICouncilConf @ianlivingstone @sentry @browserbase @Cloudflare 🫶🏻 thanks for your support it was a great conversation

[Insecure Agents Podcast @insecureagents]

Thanks to all our panelists for a great discussion on building internal AI

We learned how Cloudflare doubled MR velocity, Browserbase’s feature request pipeline runs at 100% coverage, and how Sentry turned a full debugging session into 1 slack message

@pk_iv @zeeg @ianlivingstone @irvinebroque

[NachoRulo @nachorulo_]

Great @insecureagents talk today about building internal AI agents

[Insecure Agents Podcast @insecureagents]

https://t.co/DnnGZRpZV7

[Insecure Agents Podcast @insecureagents]

TOMORROW 👀

last chance to RSVP

[Insecure Agents Podcast @insecureagents]

The rewards from shipping internal AI are clear. Take @browserbase's example:

> cut support response time to <24 hours
> feature request pipeline runs at 100% coverage
> session investigation dropped from 30–60 minutes to a single Slack message

Hear from @pk_iv & more THURSDAY

[Insecure Agents Podcast @insecureagents]

"If you care about your cognitive security, and you don't want to be manipulated, then what you should do is essentially raise your own model or use local models, local inferencing, etc." @GeoffreyHuntley

[Insecure Agents Podcast @insecureagents]

RSVP: https://t.co/DnnGZRpZV7

[Insecure Agents Podcast @insecureagents]

The rewards from shipping internal AI are clear. Take @browserbase's example:

> cut support response time to <24 hours
> feature request pipeline runs at 100% coverage
> session investigation dropped from 30–60 minutes to a single Slack message

Hear from @pk_iv & more THURSDAY

[Insecure Agents Podcast @insecureagents]

Everyone's building internal AI. A few teams are actually shipping it.

Cloudflare nearly doubled MR velocity
Browserbase cut support response to <24 hours
Sentry turned 30-min investigations into 1 Slack message

@zeeg @pk_iv @irvinebroque @ianlivingstone tell us how next week

[Insecure Agents Podcast @insecureagents]

"You gotta build as if inferencing is going to be fast. Really fast. 200 millis fast. If inferencing takes 200 millis to do, what about CICD? I see entire categories of developer tooling just gone.

But the principles of CICD and verification of software is important"

@GeoffreyHuntley

[Insecure Agents Podcast @insecureagents]

Full episode https://t.co/D020T9uGeu

[Insecure Agents Podcast @insecureagents]

"The next logical stage of a software factory is a product factory...next thing you know, you've got autonomous living software that exists to make money." @GeoffreyHuntley