AI agent safety is becoming increasingly important as agents take on more complex tasks and interact with external tools. @trishoolai (SN23) is making progress on one of the critical parts of that challenge: ensuring AI agents can safely handle their outputs. HaloGuard's output guard has now achieved a new state of the art, recording an average F1 score of 83.63 across public benchmarks, surpassing the previous best of 82.5. What makes this particularly interesting is how the team achieved these results. A few key details: • 0.8B model, roughly a quarter of the size of the previous leading model. • Designed to make continuous output monitoring more efficient and affordable. • Developed through Phase 3 on Trishool, using 7,500+ jailbreak attempts from 125+ unique miners over 15 weeks. The approach is worth paying attention to. Miners are incentivised to find weaknesses in the guard, and those failures become valuable training signals for improving the model. This creates a continuous feedback loop where decentralised red-teaming directly contributes to stronger safety systems. It also builds on HaloGuard's earlier progress with its input guard, which had already achieved SOTA performance at a similar model size. With the output guard now catching up, both sides are coming together to provide a practical safety layer for AI agents. Efficiency matters here. Safety systems need to be capable enough to identify threats without becoming too expensive or slowing down real-world deployments. This is another example of how Bittensor's incentive-driven approach can turn distributed contributions into measurable improvements in open AI infrastructure. Models and research paper are expected on Hugging Face and arXiv soon.

🔉 Exploit Summit 2026 is one week away. We’re proud to announce @SubnetSummerTAO as a livestream partner for @ExploitSummit . We’ll be streaming the event live from the Subnet Summer X account, covering talks, debates, workshops, and live demos from New City Gas in Montréal. We’ll also be on the ground hosting 2 podcasts at the event, bringing builders, subnet teams, and operators into the conversation as it happens. Whether you’re in Montréal or watching from anywhere in the world, turn on notifications and stay with us for live coverage and exclusive conversations. Break it down. Build it better. See you at Exploit.

🚨 Subnet Summer AMA x SN111 @DeSciClaims - NOW LIVE ON YOUTUBE We covered: - What SN111 Claims is and why making science machine-readable matters - The DeSci opportunity: $3 trillion of knowledge locked in human-readable papers - How miners extract claims, how validators audit for integrity - The product roadmap: turning a knowledge graph into a revenue-generating API If you're interested in decentralised science, AI-driven research infrastructure, or how Bittensor tackles knowledge accessibility, this one's for you.

🚨 Subnet Summer AMA x SN111 @DeSciClaims SN111 Claims is building the canonical claim-evidence record for science on Bittensor. It extracts scientific contributions from papers, links each claim to grounded evidence, and makes research machine-readable for AI. The subnet transforms how scientific knowledge is incentivized, validated, and accessed. 🕓 Thursday, September 17th at 4PM BST, we're hosting a live AMA with the Claims team. We'll cover: - What SN111 Claims is and why making science machine-readable matters - The DeSci opportunity: $3 trillion of knowledge locked in human-readable papers - How miners extract claims, how validators audit for integrity - The product roadmap: turning a knowledge graph into a revenue-generating API If you're interested in decentralised science, AI-driven research infrastructure, or how Bittensor tackles knowledge accessibility, this one's for you. 📍:

The panel “From Subnet to Market: Turning Bittensor Infrastructure into Real Businesses” delivered a deep and insightful discussion on building sustainable companies on top of Bittensor. Featuring leading voices from the Bittensor ecosystem: ▫️@micaelabazo – Co-Founder & CEO, Metanova Labs ▫️@0xcarro – CEO & Founder, Manifold Labs ▫️@mogmachine – Co-Founder & CEO, Taostats ▫️@gavinzaentz – Co-Founder, Leadpoet ▫️Moderated by @markjeffrey – Partner & CIO, Stillcore Capital The conversation explored the critical challenges and opportunities in turning Bittensor subnets into real-world businesses, with valuable perspectives from builders actively shipping in the ecosystem. Video highlights from the session▶️