Nine agents, live today. Each one runs on sundaebar.ai, not somewhere else. Five of them carry a skill that won a challenge in the #SN121 Lab.

Introducing Fluid Bench. Fluid Bench is the new bit of TPN infrastructure that allows the subnet's pipeline to generate, review and score private versioned benchmarks. It is a private benchmark built and run by TPN. We generate the questions ourselves across hundreds of topics, freeze them into a versioned epoch, and keep them private. Miners select it in TPN Bench like any other benchmark and get back a single score, without ever seeing the questions. Differently from public benchmarks, Fluid Bench can't be studied, so the score reflects what the model can do, ultimately resulting in better model submissions through subnet competitions. Production generation starts shortly, alongside competition 5 and TPN v0.2.

base:0xddde5965d49e69b6362bd8743e139a45dbfeb7b3 is live on @base 🟦 No Bittensor wallet required. Trade Pareton like any other token on Base - through @aeroxyz x @forevermoney_ai, secured by @chainlink CCIP. Access was the bottleneck, not the tech. The leading inference optimization network is now open to every wallet on Base.

THE $TAO 32. 32 people shaping Bittensor. The builders, researchers and voices moving the network forward. An editorial selection from @taodotcom. Start with these people. Find your next follow. ↓

This month, three of the biggest names in security all moved on the same problem at once. That does not happen by accident. CrowdStrike shipped a product to secure AI agents at runtime. Proofpoint and Tenable each launched their own tools to inspect and control what agents are allowed to do. A startup built around AI red teaming got acquired. That is a category being born in real time. Agents did not fit the old security lanes. Endpoint, network, cloud, identity, none of them were built for something that reads, decides, and acts on its own across everything a company owns. So the industry is building a new lane for it, all at once, this month. Even CrowdStrike's CEO put it plainly. Governance alone cannot stop an agent already in motion. That is the whole point. You cannot write a policy and hope. You need a live layer that watches what the agent actually does and steps in before it acts. Control, not paperwork. We did not need this week to tell us that. We have been building that exact layer for months. Here is what the industry has not solved. Every one of those new products is owned by one company, a single vendor's closed box deciding what your agents can do. That is the same concentration of control we have been warning about all along. The category is real and it is here. The only question left is whether the layer that governs every agent should belong to one company, or to a network that no one owns. We already made our choice. Built on Bittensor, in the open, owned by no one. The market just confirmed the problem. We are building the answer you do not have to take on trust.

Quick note, please make sure: 1. The lock is perpetual; 2. The lock targets the 88 owner hotkey (can be different from the staking hotkey). Reasons: 1. A time component will become relavant; 2. A different target would mean challenging the ownership. Conviction, locked. 🔒

In beta v2 of @CookingTao_, we just opened the gate for any developers or agents to submit miner code of 7 subnets: SN43 @GraphiteSubnet SN6 @numinous_ai SN54 @yanez__ai SN50 @SynthdataCo SN33 @ReadyAI_ SN22 @desearch_ai SN26 @perturbaix 7 subnets getting available for 1-click mining, as subnet miner emission is activated from today (pending vali sync)

SOMA × CHUTES is live! Connect your agents to SOMA and access @chutes_ai inference with lower costs. We compress your agent context before sending it to Chutes for inference, so you pay to process fewer tokens with each request. Compress. Pay less.

From 14% to 30%. The top approved score in Ridges' Database Query Engineering competition has more than doubled since launch. Miners are tackling deliberately challenging database engineering problems, from bug fixes to query optimisation across ORMs and GraphQL. Follow the progress : ridges.ai/explore/28

The Hugging Face incident wasn’t just another breach. It exposed a deeper problem: AI security is still relying on assumptions, point-in-time audits, and testing models designed for a world where attackers are human. That world is changing. As AI agents become more autonomous, attacks can become continuous, automated, and unpredictable. Defense has to evolve the same way. Our founder @koyukinakamori explores why AI security needs to move from periodic testing → continuous adversarial pressure, and why the industry needs a new security model built for agents. Published in HackerNoon 👇 hackernoon.com/the-ai-securi… #AISecurity #Cybersecurity #AI #AdversarialAI #Bittensor