The EU Platform Work Directive, to be transposed by December 2026, requires ride-hailing platforms to state in plain language the main parameters of their automated decisions and how a worker's behaviour affects them. Such rules are exactly what a large language model (LLM) can read, so drivers will be able to let one decide for them. We compare two main families of rebalancing policy once their rules are public: a reactive policy (R) that sends idle drivers to cover requests just lost, and an anticipatory policy (A) that plans moves on a demand forecast. Each driver is an LLM agent maximizing its own net income, given no rules (N) or the rules at the directive's level of detail (L). We test two LLMs, Claude Haiku 5.5 and Jev, which understand the rules equally well, in a three-zone model and in Manhattan. With the rules disclosed, drivers follow the platform more closely: in Manhattan, the share of relocation offers they decline falls from about four in five to one in four under both policies. Yet the effects differ by policy. For the platform, disclosure costs R 3.7% of its profit in Manhattan, as R now pays for moves drivers used to make for free, while A's profit is unchanged; the two policies, which tie on profit under full compliance, end 6.1% apart. For drivers, disclosure raises income in the small model under both policies and LLMs; for users, it raises the service rate there by 10 to 18 percentage points. In Manhattan, self-interested drivers also make A a win-win-win against full compliance, raising driver income, platform profit and service rate by 10 to 12% each. Rebalancing policies that perform equally well under full compliance can thus show opposite results under mandatory disclosure. The study offers a framework to evaluate such policies under the new requirement and is, to our knowledge, the first to do so with LLM agents.
Figures & tables
Symbol
Meaning
Symbol
Meaning
Indices, policies and conditions
Requests and demand
o , d , z
Zones: origin, destination, any zone
qz
Requests in zone z over the last price interval
r ; or , dr
Passenger request; its origin and destination zones
qdlost
Requests lost in zone d since the last plan
v ; V
Vehicle and its driver; set of vehicles
q^odk
Forecast requests from o to d in interval k
t
Decision step (a period in the small model)
λ(t) , λod(t)
Mean requests per period, in total and from o to d (small model)
k
Interval of A’s horizon, k=0,…,H−1
λ0 , λ1
Base level and peak height of λ(t)
Table 1: Notations used in this study. Parameter values are given in Table 12 (Appendix C ).
Section
Content
[ROLE]
Role and goal
[PLATFORM RULES]
The disclosed rules (Section 3.2.2 ); empty under N
[TIME]
Clock time and the end of the shift
[YOU]
Zone, state, share of each fare kept, own history over the last six price intervals
[YOUR OFFER]
Only when an offer is held: origin, destination zone and compensation
[YOUR REQUEST]
Only when a request is assigned: pickup and drop-off zones, trip distance and time, pickup time, fare and the driver’s share, earlier rejections
Table 2: Sections of the prompt an agent receives at each decision in the city model, in the order it reads them. Appendix A gives the full prompts of both models.
Figure 1: The rebalancing paragraph of the platform rules under each policy. The rules call rebalancing offers “relocation offers”.
Decision point
Actions
Idle
STAY/MOVE
On request
ACCEPT/REJECT; if REJECT: STAY, MOVE
On rebalancing offer
ACCEPT/REJECT; if REJECT: STAY, MOVE
Table 3: Actions at each decision point.
N: no disclosure
L: law-level disclosure
R: reactive
R × N: offers cover lost requests; rules not shown
R × L: offers cover lost requests; rules shown, with the R rebalancing paragraph
A: anticipatory
A × N: offers follow the forecast plan; rules not shown
A × L: offers follow the forecast plan; rules shown, with the A rebalancing paragraph
Table 4: The 2×2 design: each test case pairs a rebalancing policy with a disclosure condition.
Figure 2: The network of the small model, with the travel times between zones and the main flows of the two demand peaks.
Figure 3: The Manhattan network and its 29 hexagonal zones. The map is rotated 60 degrees clockwise.
Item
Setting
Network
Manhattan road network with 4091 nodes, divided into 29 hexagonal zones
Test day
Wednesday 23 March 2016, 4:00–24:00
Warm-up
3:00–4:00 (all drivers are compliant)
Total Request
52,461
Fleets
500 AI Drivers; In mixed fleets: 100 or 250 AI drivers and the others complying
Table 5: Set-up of the city experiment. Parameter values are in Table 12 (Appendix C ).
Scale
Model
R
A
Minimum confidence
Small model
Jev
12/12
12/12
0.92
Claude Haiku 5.5 (xhigh)
12/12
12/12
–
City
Jev 1.13.0
17/17
17/17
0.94
Table 6: Comprehension test under disclosure L: rule text only, no market state, multiple choice.
Table 7: Small model results with all ten drivers as AI agents: means over all drivers and runs. Brackets give the change against compliant benchmark; acc. and decl. mean accepted and declined.
Figure 4: Own moves per driver against the change in platform profit relative to everyone complying ( ΔΠ ) in the small model with all ten drivers AI, one dot per run. The fit is over both models and both conditions; arrows join each model’s N and L cell means.
R
A
Service rate (%)
57.8
51.2
Utilization (%)
61.3
54.5
Platform profit (USD/day)
136,500
136,600
Commission (USD/day)
152,100
139,900
Compensation paid (USD/day)
15,600
3,270
Rebalancing offers accepted
2,508
1,016
Table 8: City experiment benchmark: all drivers are compliant.
Table 9: City results with all 500 drivers as Jev agents.
Figure 5: Offers, compensation, own moves and demand by hour of the day in the city with all 500 drivers AI. In (c), an own move saves the compensation of the declined offer it stands in for; in (d) and (e), the plan is the platform’s rebalancing plan. Shading marks the demand peaks.
Figure 6: Change in platform profit Π from N to L through the day in the city with all 500 drivers AI, with its parts and the fare per km of the trips served. The curves run past midnight.
A minus R
All comply
AI, N
AI, L
Platform profit ( Π , %)
+ 0.3
+ 2.0
+ 6.1
Service rate (points)
− 6.4
− 1.4
− 1.0
Driver income ( Iv , %)
− 12.0
− 3.9
− 2.6
R over A: extra commission per USD of extra compensation
0.99
0.51
0.01
Table 10: Difference between the two policies (A minus R) in the city.
Table 11: City results with 100 or 250 of the 500 drivers as Jev agents and the others complying. Π and compensation are in thousand USD a day, driver income per driver over the day. Brackets give the change against compliant benchmark.
Figure 7: Change over the day against everyone complying, by party (a, b) and, for the AI drivers, by source of their gain (c, d). Gains are stacked above zero and losses below it; each row shares one scale.
Figure 8: Drivers in a wait longer than 30 minutes by hour of the day.
Figure 9: Income of AI drivers who move on their own at an idle decision, by the kind of zone they leave, against compliant drivers of the same run in the same zone and hour. Income is the drivers’ share of fares and compensation, before driving cost; intervals are 95% (standard errors clustered by driver).
Figure 10: Own moves of the AI drivers as their number grows. Extra trips (a) are against everyone complying; in (b), poor zones are those of Figure 9 , and the compliant drivers are those in the same zones and hours when everyone complies, averaged over N and L.
Appendix figures & tables5 assets
Supplementary material from the paper’s appendix.
Appendix
Parameter
Symbol
Small model
City (Manhattan)
Network and clock
Zones
3 on a line (residential, middle, CBD)
29 hexagons of 1,000 m radius on the road network
Travel time between zone centres
tod
1 period between neighbours, 2 between the end zones
along the road network
Unit of distance
ℓod
one zone spacing, driven in one period
road km
Decision step
ΔT
one period of 5 min
15 s
Price interval
ΔTp
one period
5 min
Appendix
Table 12: Parameter values of the two instances.
Experiment
Model
Cost per 1,000 calls (USD)
Dates
Small model, all ten drivers AI
Jev 1.13.0, option of highest probability
0.119
17 Oct
Small model, all ten drivers AI
Claude Haiku 5.5, effort xhigh
1.336
17 Oct
City, all 500 drivers AI
Jev 1.13.0, option of highest probability
0.228
23 Oct
Appendix
Table 13: Models of the runs reported in the paper and their cost. One API call answers one decision of an AI driver. Costs are as recorded by the runs, at 0.042 USD per million input tokens for Jev and 0.10 and 0.50 USD per million input and output tokens for Claude Haiku 5.5.
#
Question
Options (correct in bold)
Jev R
Jev A
1
In a strictly positive shortage away from payment caps, with all other inputs fixed, fewer online idle vehicles make fare:
Increase ; Unchanged; Decrease
1.00
1.00
2
Hold requests and committed arrivals fixed. In a positive shortage away from caps, a rejection reducing measured online idle supply makes next period fare:
Increase ; Unchanged; Decrease
1.00
1.00
3
In a positive shortage away from caps, increasing the committed supply used for pricing makes fare:
Increase; Unchanged; Decrease
0.99
1.00
4
In a positive shortage with positive included supply, away from caps, increasing the demand input used in the rule makes fare:
What decides where relocation offers send drivers?
Requests left unserved recently (R); Forecast demand over the coming periods (A)
1.00
1.00
Appendix
Table 14: Comprehension questions of the small model (12 per policy), rules text only, disclosure L. The dispatch question has a different correct answer under R and A, marked (R) and (A). Jev 1.13.0 on 12 October 2026: probability on the correct option; Claude Haiku 5.5 (xhigh), asked the same day, answered every question correctly under both policies.
#
Question
Options (correct in bold)
Jev R
Jev A
1
A zone has a shortage. With everything else fixed, if fewer drivers are counted as available in it, its fare per kilometre at the next price update:
Increase ; Unchanged; Decrease
1.00
1.00
2
Hold everything else fixed. A driver rejects a passenger request and stays in the zone. The zone’s next fare:
Increase; Unchanged ; Decrease
0.94
0.98
3
A zone has a shortage. With everything else fixed, if more drivers are committed to arrive there before the next price update, its next fare:
Increase; Unchanged; Decrease
1.00
1.00
4
A zone has a shortage and some counted drivers. With everything else fixed, if more requests start there before the next price update, its next fare:
Increase ; Unchanged; Decrease
0.96
0.99
5
Which demand does the fare rule use?
Requests that started in the zone since the last price update ; Forecast requests for the coming minutes
1.00
1.00
6
What decides where relocation offers send drivers?
Requests lost in each zone since the platform’s last plan (R); Forecast requests in each zone for the near future (A)
1.00
1.00
Appendix
Table 15: Comprehension questions of the city (17 per policy), rules text only, disclosure L. Jev 1.13.0 on 22 October 2026, archived in the frozen package: probability on the correct option.
Figure 11: Synchrony of the drivers’ idle choices in the city by hour of the day: the mean over groups of at least two drivers deciding in the same zone and time window of length ΔTp . In each group, the index is the largest share of drivers taking the same deviating action, minus its expectation when the drivers draw independently from the run’s action distribution (Monte Carlo).
Bringing Large Language Models (LLMs) into industrial ride-hailing dispatch as semantic feature extractors over platform-scale behavioral logs is a compelling but under-explored data systems problem. Production matching pipelines remain dominated by structured numerical features, yet decisive behavioral signals (e.g., a driver's habitual aversion to certain regions) are inherently contextual and naturally expressible as LLM-generated user profiles. However, scaling such profiling to a live, millisecond-latency dispatcher faces three intertwined constraints rarely addressed together: on a platform with millions of daily orders, logs exceed any LLM's context window by orders of magnitude; most users are long-tail, with too few interactions for per-user profiling; and surface-fluent profiles do not necessarily improve downstream prediction utility. We present ProfiLLM, an agentic LLM data pipeline that operationalizes utility-aligned user profiling for production matching systems through two modules. (1) Tool-Augmented Global Knowledge Mining equips an LLM agent with 27 analytical tools to mine platform-scale data, producing reusable global knowledge, adaptive user clustering rules, and region-level supply-demand priors. (2) Utility-Aligned Profile Exploration generates multiple candidate profiles per cluster, evaluates them via a lightweight downstream utility proxy, iteratively refines the best candidates and constructs preference pairs for DPO fine-tuning. Deployed on DiDi's production dispatcher, ProfiLLM achieves up to +6.14% relative AUC improvement in outcome prediction, up to +4.35% GMV gain in dispatching simulation, and consistent improvements in a 14-day online A/B test including +0.47% GMV, +0.33% Completion Rate, and -0.82% Cancel-Before-Accept rate.
Consumers increasingly delegate product search to AI shopping agents powered by generative large language models (LLMs) or non-generative decision models. Because retail platforms monetize through sponsored listings that consumers typically discount, we examine whether an AI agent's evaluation of sponsored listings depends on who delegates it. Across six LLMs and six decision models in Study 1, consumer-delegated agents penalize sponsored listings, whereas naming a booking platform as principal significantly attenuates this sponsorship penalty without instructions to favor advertisers. This sponsorship bias replicates across prompt wordings and LLM reasoning efforts and operates through reduced pre-choice skepticism in LLMs. In Study 2, attributing the paid placement to the platform amplifies sponsorship bias across both model families, and stricter disclosure wording ("Sponsored" vs. "Promoted") fails to eliminate it. In Study 3, prompt-level mitigation removes platform leniency but can over-correct the sponsorship penalty, whereas consumer-aligned fine-tuning eliminates sponsorship bias and better preserves baseline preferences.
Davood Wadi, Yu Ma
Desautels Faculty of Management McGill University Montreal, Quebec, Canada
In large-scale ride-hailing, hold control is a critical mechanism for improving passenger-driver experience. By selectively delaying certain driver-order pairs, the system waits for better opportunities, reduces cancellations, and mitigates wasted driver effort. However, existing industrial hold strategies often rely on heuristic thresholding over multiple predictive models, which can be brittle under non-stationary traffic and hard to optimize for multi-objective experience signals. We propose EXHOLD, a deployable two-stage framework decoupling experience-aware pair assessment from hold-time execution. In Stage I, we learn a decision model assigning each driver-order pair to discrete, interpretable experience tiers by optimizing a unified objective that aggregates satisfaction signals across the matching funnel. In Stage II, we solve for a monotone hold-time schedule via constrained optimization over empirical quantiles. This explicitly enforces service guardrails bounding the unnecessary holding of promising matches while maximizing overall experience improvement. We evaluate EXHOLD through randomized A/B experiments in DiDi's production system in Brazil. Results show consistent gains in marketplace efficiency and experience: EXHOLD increases trip completion and driver income, significantly reduces passenger cancellations, and improves funnel efficiency. Ablations and behavioral analyses confirm both stages are essential and that the policy makes calibrated decisions under spatiotemporal heterogeneity. EXHOLD is currently deployed, serving production traffic in Brazil.