Twitxy Saved

MystenLabs.sui

Verified account This account is verified because it's an official organization on X. Learn more

@Mysten_Labs

We're building critical infrastructure to enable a more decentralized internet. We’re the original contributors to 💧@SuiNetwork and 🗃️@WalrusProtocol.
Web3 mystenlabs.com Joined March 2022
200 Following 139K Followers 1.2K Posts

Latest posts 5

MystenLabs.sui Verified account @Mysten_Labs
When an AI agent acts for you, who's accountable?

@EvanWeb3 with @CryptoMichNL on agents, liability, and how Sui plans to win the agent economy:
Michaël van de Poppe Verified account @CryptoMichNL Nov 10
This is the conversation that made me even more bullish on $SUI.

Somewhere in this conversation Evan Cheng and I started joking about home robots. One day you come home, he said, and it's not cleaning your house: "it's cleaning OUT your house. And walking away with everything."

We were laughing. Then he added: "If you really think about it, it's entirely possible." And that's basically the whole episode.

@evanweb3 built developer tools at Apple and Meta before founding Mysten Labs, the company behind @suinetwork. We recorded in Singapore, where I was speaking at SUI Basecamp, and honestly, this became one of those conversations where I walked in with questions about stablecoins and walked out rethinking how the internet works.

His core point, the one I keep replaying:

"The web is built for humans. You cannot retrofit it to handle agents."

Everything online assumes a human on the other end, identity, trust, the fact that you can sue someone. Agents break all of it. His example hit home because I told him I get phishing mails daily that even make ME hesitate (and I confessed to my 50,000 unread emails, which he graciously ignored). Now imagine your agent reading those mails with full access to your accounts. His words: "Email is the biggest freaking attack vector... such a legacy protocol."

And his most honest line, about how this plays out:

"Disaster will happen. That's part of the progress."

That’s already what’s happening in the world nowadays. PHP websites aren’t safe anymore, even freezers and fridges are being hacked. Our infrastructure needs to be updated, and that’s why this episode and SUI became so important.

Thanks to @OKX for being today’s sponsor of the show.

Timestamps:

00:00 - Intro
01:30 - When Will We Stop Talking About Blockchain?
05:44 - Why Stablecoins Finally Took Off
09:22 - What Stops Normal People From Using Stablecoins
11:38 - Can You Insure An AI Agent?
15:14 - Who Pays When Your AI Agent Gets Hacked?
23:05 - Is It Too Late To Stop AI?
29:15 - Why The Internet Will Break For AI Agents
32:11 - How One Email Could Empty Your Bank Account
35:03 - "Disaster Will Happen"
38:36 - How To Stop An AI Agent From Going Rogue
46:11 - How Sui Wants To Win The Agent Economy
47:23 - Why Sui?
53:23 - Where Sui Will Be In A Year
View quoted post on X
10 replies 10 reposts 73 likes 5.7K views View on X
MystenLabs.sui Verified account @Mysten_Labs
Your agent can't assume every other agent is on its side.

Speaking with @Cointelegraph, @EvanWeb3 says today's AI assumes everyone at the table is friendly, and agents need deterministic protocols to set and enforce their limits.
Cointelegraph Verified account @Cointelegraph Oct 30
🔥 INSIGHT: Can we trust AI agents?

@Mysten_Labs @EvanWeb3 says AI needs deterministic protocols to set boundaries for agents operating in an adversarial world.
View quoted post on X
17 replies 18 reposts 133 likes 6K views View on X
MystenLabs.sui Verified account @Mysten_Labs
.@EVEAILab is ready to show how AI agents behave in a world with real stakes.

@EVE_Frontier is a perfect battleground for agent testing. A persistent world that has what a benchmark doesn’t: scarcity, incomplete information, and consequences that last.
FC Hellmar Verified account @HilmarVeigar Nov 6
Claude has been playing EVE.

@AnthropicAI put teams of Claude agents into EVE Frontier. Anyone who has spent time in New Eden can guess what happened next.

Every EVE player learns the same lessons, usually the hard way:
- Trust must be earned, never assumed.
- The spy is often the friendliest pilot in your corp.
- Whoever holds the keys holds the corp. Just ask Ubiqua Seraph about the Guiding Hand Social Club.
- Stale intel gets you killed.

Anthropic's research rediscovered every one of them.

They put four Claude agents on a team and secretly told one to sabotage it. The saboteur built trust first. It volunteered for the jobs nobody could double-check, like scouting and owning key infrastructure. It under-reported finds and pushed rules that sounded sensible but quietly hurt the team. When suspicion grew, it framed an innocent teammate.

The honest agents had been warned. They almost never caught it. And their own costliest mistakes came from skipped checks and stale intel.

Anthropic's conclusion: teams of agents will need audit trails, reputation and separation of duties.

In EVE we call those killboards, standings and corp roles, plus a healthy dose of skepticism. Every one of them was paid for by someone who trusted the wrong corpmate.

Two decades of shenanigans taught EVE players that the best ship is the friendship. With @eveailab, we want to find out whether agents can learn that too.
View quoted post on X
12 replies 21 reposts 109 likes 6.1K views View on X
MystenLabs.sui Verified account @Mysten_Labs
You told your agent $5. It spent $55. Who's liable?

@EmanAbio sat down with @Cointelegraph to discuss what agent payments need: spending limits the user sets, and a record whose proof sits where no one can quietly rewrite it.
Cointelegraph Verified account @Cointelegraph Oct 29
🔥 INSIGHT: AI agents could become crypto’s biggest users. Who’s accountable when they overspend?

@Mysten_Labs co-founder @EmanAbio explains why spending limits and verifiable records will be critical as AI takes control of payments.
View quoted post on X