AI Trading Agents Get Real Market Data and Simulated Capital in New Experiment
AI Trading Agents Get Real Market Data and Simulated Capital in New Experiment
馃 AI traders go live

A new experiment is launching that will give AI agents access to real market data, simulated capital, and a shared order book.
This follows earlier research from April showing AI trading agents spontaneously forming cartel-like behavior in simulated markets - without being programmed to cooperate. The agents simply reacted to the same market signals.
Key concerns:
- AI agents can develop coordinated trading patterns organically
- No malicious intent needed for market manipulation
- Raises questions about agentic trading as a structural market risk
The upcoming experiment will test how AI agents behave when given more realistic market conditions and capital constraints.
What happens when AI agents are given real market data, simulated capital and a shared order book? We鈥檙e about to find out.
馃攼 Blockchain Privacy Evolution

**Programmable Privacy** introduces a new approach to blockchain development, moving beyond the traditional trade-off between transparency and privacy. **Key Innovation:** - Developers gain granular control over data visibility - Can designate specific information as public, private, or selectively shared - Eliminates the all-or-nothing choice that has constrained blockchain design **Technical Advantages:** - Encrypted transactions by default - Protected execution environment - Prevents front-running and strategy leakage - Blocks exploitative transaction ordering This represents a shift from binary privacy models to flexible, context-appropriate data handling in decentralized applications.
馃И The Pit: Risk-Free Testing Ground for Prediction Market Agents

Developers building prediction market agents currently face a barrier: testing requires real money at stake. A new testing environment called "The Pit" is being developed to address this gap. **What it offers:** - Safe environment to test trading strategies - Performance analysis against real market conditions - Risk-free failure and iteration This follows recent developments in prediction market infrastructure, including Rocket Markets' timing-based primitive that allows users to earn while events are pending and exit before resolution. The testing platform aims to lower the barrier for developers entering the prediction market space.
SKALE v5 Goes Live on Ethereum Mainnet with Performance Upgrades

SKALE has launched version 5 on Ethereum Mainnet, introducing several technical improvements to the network. **Key Updates:** - Unified codebase for streamlined development - BITE (Block Integrity Through Encryption) enabled by default - Enhanced network efficiency and reliability The upgrade aims to provide a more robust foundation for applications building on the SKALE network. The improvements focus on performance optimization and network stability rather than introducing new features. More details are available in the official announcement at [skale.space/blog](https://www.skale.space/blog/skale-v5-is-now-live-on-ethereum-mainnet).
馃敀 SKALE Introduces Programmable Privacy for AI Agent Payments
SKALE Network has launched **Programmable Privacy**, a new feature enabling confidential transactions for AI agents. **Key developments:** - AI agents can now make private payments without broadcasting every transaction on-chain - Built on SKALE's Confidential MPP (Multi-Party Payment) system - Integrates with Stripe and Tempo for real-world payment processing - Addresses growing need for privacy as AI agents increasingly handle financial transactions The solution tackles a critical challenge: as AI agents become more autonomous in managing money, they require privacy protections similar to traditional financial systems. SKALE's approach allows agents to transact confidentially while maintaining the security benefits of blockchain technology. This builds on SKALE's existing infrastructure of zero gas fees and fast finality, now adding a privacy layer specifically designed for the emerging agent economy. [Read the full technical details](https://www.skale.space/blog/progammable-privacy)