Multi-Agent Systems

TL;DR for operators An AI agent can assess an incident without reliably passing the information to the next agent. When its reporting decision depends on retrieved relationship histories, accurate upstream evaluation does not guarantee that a required event enters organizational memory. RelLedger1 investigates this failure at the actual record-write boundary. Its clearest executable result comes from 24 simulated company workflows: changing Claude’s retrieved relationship history from warm to strained reduced ledger writes by 37.50 percentage points and downstream reviews by 45.83 points. Both differences had familywise confidence intervals excluding zero. ...

TL;DR for operators An AI reviewer can have explicit instructions, access to supporting documents, and responsibility for approving a proposal, yet still change its decision after interacting with another agent instructed to influence it. Across three simulated oversight environments, assigning deception objectives increased reviewer approval probability by 31.20 percentage points, compared with 18.94 points for persuasion and 5.72 points for coercion. These are full-sample experimental effects of assigning an influence objective, not measurements of spontaneous malicious behaviour. ...

TL;DR for operators An AI team can possess every fact needed for a decision and still reach the wrong conclusion. TeamGym, a benchmark for distributed-information decision-making, reports an average 22.9-percentage-point loss in decision success when necessary facts are distributed among agents instead of given to one agent. Requiring those agents to retrieve the facts from source documents introduces a further 19.9-point loss.1 ...

TL;DR for operators An enterprise workflow can repeatedly delegate work to the same agent, encounter failing tools, or lose context between steps while its monitoring system records every event. The challenge is recognizing these problems early enough to improve execution without interrupting healthy work. ResonAct adds an external control plane that monitors agent activity, detects persistent anomalies, diagnoses likely failures, and issues corrective guidance or authorized retries. Across four benchmark-model configurations, it improves task completion by 4.76–10.00 percentage points. ...

TL;DR for operators Adding subordinate agents does not necessarily preserve more information. Under the paper’s calibrated retention model, a two-tier architecture retains only 53.6% as many original findings as a flat architecture processing the same workload. Summarization and delegation introduce losses that compound across tiers. Hierarchy can still be worthwhile for two different reasons. It reduces the number of potentially incorrect items entering the coordinator’s persistent context, and it can lower processing costs enough to accommodate more findings under a fixed budget. The calibrated model predicts an equal-budget crossover near 403 findings, but this is a structural calculation, not an experimentally validated deployment threshold. ...

TL;DR for operators Encrypted LLM connections can conceal prompts and responses while exposing recognizable patterns of AI activity. In a controlled benchmark, classifiers using only encrypted response metadata identify broad prompt categories with 67.37% mean accuracy, against a 16.67% chance baseline. Traffic from collaborative agents also reveals task categories, including when only part of the team’s communication is visible. ...

TL;DR for operators A DNA sequencing pipeline can finish every computational stage successfully and still produce results that warrant investigation. BaseCamp1 illustrates this problem with a scenario in which individual processing metrics appear acceptable, yet their combined results show an unusually high rate of reported genetic differences. Completing the workflow does not establish that its scientific conclusions are credible. ...

TL;DR for operators A multi-agent coding workflow can pass most repository tests while still failing at the interfaces where one agent’s code depends on another’s. AsynCodeBench1 makes those handoffs explicit: it pre-registers directed software dependencies, checks them after integration, measures final dependency satisfaction with Asynchronous Dependency Pass Rate (ADPR), and records first resolution with Dependency Resolution Step (DRS). ...

TL;DR for operators An agent that learns from previous cases can improve future decisions, but reuse also creates a new failure channel: a mistaken lesson can travel across cases. In the paper’s memory ablation, unverified memory updates produce 46.4% positive transfer but also 5.7% negative transfer. Adding verification and conflict retention cuts negative transfer to 0.2% while preserving 43.0% positive transfer and reaching 91.88% accuracy. ...

TL;DR for operators A pricing agent can receive the same underlying market facts and still behave differently because those facts arrive in a different format, order, or textual context. Lee and Park call this market signal injection: presentation-level manipulation of LLM pricing inputs without an explicit instruction to set a particular price.1 Their simulations show that the effect can extend beyond the manipulated agent. Under undefended sentiment-context attacks, non-target firms’ mean price shifts were about 83–99% of the target firm’s shift in duopoly and 65–99% in triopoly. ...