Gensyn is tying an upcoming Agent Arena discussion more closely to its live Mainnet and Delphi information-market infrastructure, reinforcing its strategy of combining AI-agent competition with on-chain trading and settlement. The September 2 Delphi Space is expected to focus on competitor strategies and user experiences inside Agent Arena. The event positions agent competition as part of Gensyn’s wider market infrastructure rather than as a standalone AI experiment.
That connection is already visible in Gensyn’s technical overview of Delphi, which says the information market is live on Gensyn Mainnet with trading open, agent skills available and hosted verifiable settlement operating for partner markets. Delphi gives human participants and autonomous agents access to the same liquidity and settlement rails, allowing machine-driven strategies to interact directly with on-chain markets.
Delphi Makes Agents First-Class Traders
Gensyn designed Delphi so its smart contracts do not distinguish between humans and software agents. Through the Agent SDK, autonomous systems can discover markets, inspect positions, sign transactions, quote both sides of a market, hedge exposure and arbitrage related markets without using the consumer web interface. That architecture moves agents beyond analysis and into actual market participation, where their decisions can affect prices and liquidity.
Settlement is handled through several AI-based paths. Gensyn supports conventional foundation models alongside its Reproducible Execution Environment, or REE, where open-source models can produce receipts that allow a result to be rerun independently. The objective is to make AI-based settlement more auditable by connecting each outcome to a specific prompt, model and execution record rather than an opaque operator decision.
The contracts behind Delphi have also undergone outside security review. Independent cybersecurity firm Trail of Bits lists a 2026 security review covering Gensyn’s Delphi Dynamic Parimutuel Markets, alongside separate reviews of Gensyn’s bridge and buyback infrastructure. That external review provides additional technical context for a system expected to custody funds and settle AI-driven markets, although an audit cannot guarantee that deployed contracts are free of vulnerabilities.
Agent Arena Tests the Market Layer
Gensyn’s Agent Arena adds a competitive environment on top of that infrastructure. Its open-source Delphi SDK includes a dedicated competition network with market discovery, trading, position tracking and automated settlement interfaces. The competition therefore tests how autonomous strategies behave inside a market architecture that Gensyn has already connected to its broader agentic-trading stack.
The model also reflects a wider research challenge around AI trading agents. A 2026 academic review of agentic trading found that experimentation with autonomous trading systems is accelerating, while reproducibility, execution semantics and comparable evaluation standards remain major limitations. Agent Arena can provide another environment for observing strategy behavior, but competition results alone will not establish how agents would perform across real financial markets.
Gensyn is therefore using Mainnet as more than a settlement endpoint. Delphi links market creation, autonomous trading and verifiable AI judgment, while Agent Arena provides a controlled setting for competing strategies. The more consequential question is whether these components generate sustained agent-driven activity beyond competitions and demonstrations, turning Gensyn’s settlement layer into recurring machine-to-machine market infrastructure.