analiza aiqbench
This is a measurement study of LLM-driven options-trading agents, scoring 26 systems on whether they enforce safeguards outside the model and how large a loss their order paths can reach. The headline finding is that no system authenticates its text sources or cross-checks market data, and 13 can reach unbounded loss. For AI readers, the key point is that prompt-level manipulation can change an order's strike, expiry, or legs while its direction stays correct, so loss limits need to sit at the order boundary rather than rely on the model.
Unbounded Losses, Unchecked Inputs: Measuring the Safeguards of LLM Options-Trading Agents
arXiv:2610.24289v1 Announce Type: cross Abstract: Large language model (LLM) agents now place options orders at retail brokers, and developers publish open-source options agents faster than anyone has examined them. Prior security work on trading agents scored academic equities designs by whether an attacker can flip a buy-or-sell decision. We present the first measurement of the options agents that brokers and developers ship, scored by the safeguards they enforce and by the worst loss their documented order paths can reach. We screened 765 GitHub repositories, broker documentation, and the research literature and found 215 eligible repositories and 8 broker-native agent interfaces. We then scored 26 systems, including every broker interface and every open-source project with at least 20 stars, on seven robustness and eleven security safeguards with cell-level evidence. The ecosystem is new: 198 of the 215 repositories were created in 2026, and 129 (60.0%) appeared in the two weeks around one options-trading hackathon. Of the 22 systems that can place orders, 10 enforce a pre-trade limit outside the model, yet none of the 26 authenticate their text sources or cross-check market data. Of the 26, 17 can execute with no mandatory human step and 13 can reach unbounded loss, including 6 of the 8 broker interfaces. Because an attacker who writes the agent's inputs can move an order's strike, expiry, volatility, and legs while its direction stays correct, options agents need loss limits enforced at the order boundary.
Abstract:Large language model (LLM) agents now place options orders at retail brokers, and developers publish open-source options agents faster than anyone has examined them. Prior security work on trading agents scored academic equities designs by whether an attacker can flip a buy-or-sell decision. We present the first measurement of the options agents that brokers and developers ship, scored by the safeguards they enforce and by the worst loss their documented order paths can reach. We screened 765 GitHub repositories, broker documentation, and the research literature and found 215 eligible repositories and 8 broker-native agent interfaces. We then scored 26 systems, including every broker interface and every open-source project with at least 20 stars, on seven robustness and eleven security safeguards with cell-level evidence. The ecosystem is new: 198 of the 215 repositories were created in 2026, and 129 (60.0%) appeared in the two weeks around one options-trading hackathon. Of the 22 systems that can place orders, 10 enforce a pre-trade limit outside the model, yet none of the 26 authenticate their text sources or cross-check market data. Of the 26, 17 can execute with no mandatory human step and 13 can reach unbounded loss, including 6 of the 8 broker interfaces. Because an attacker who writes the agent's inputs can move an order's strike, expiry, volatility, and legs while its direction stays correct, options agents need loss limits enforced at the order boundary.
26 pages, 9 figures, 8 tables. Original measurement study: a logged screen of 765 GitHub repositories and cell-level scoring of 26 LLM options-trading systems against 18 safeguards
Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI)
arXiv:2610.24289 [cs.CR]
(or arXiv:2610.24289v1 [cs.CR] for this version)
Text integral sub licența deschisă a publicației · 275 cuvinte