Perly Consulting │ Beck Eco

The State of Play

A living index of AI adoption across industries — where established practice meets the bleeding edge
UPDATED DAILY

The AI landscape doesn't move in one direction — it lurches. Some techniques leap from experiment to table stakes in a single quarter; others stall against regulatory walls, technical ceilings, or organisational inertia that no amount of hype can dislodge. Knowing which is which is the hard part. The State of Play cuts through the noise with a rigorously maintained index of AI techniques across every major business domain — classified by maturity, evidenced by real-world adoption, and updated daily so you always know where you stand relative to the field. Stop guessing. Start knowing.

The Daily Dispatch

A daily newsletter distilling the past two weeks of movement in a domain or two — delivered to your inbox while the index updates in the background.

AI Maturity by Domain

Each dot marks the weighted maturity of practices within a domain — hover for a brief summary, click for more detail

DOMAIN
BLEEDING EDGEESTABLISHED

Collections, payments & purchase order automation

GOOD PRACTICE

TRAJECTORY

Stalled

AI that optimises collections, predicts payment timing, and automates purchase order matching and processing. Includes dunning optimisation and three-way PO matching; distinct from invoice processing which handles incoming vendor bills rather than outgoing payments and procurement.

OVERVIEW

Automating collections, payments, and purchase order workflows is a proven, GA-tier practice with documented vendor momentum, quantified ROI, and analyst validation — the question has shifted from "does it work" to "how fast can enterprises implement and at what organizational cost." Agentic AI infrastructure now enables autonomous negotiation, intelligent cash application, and compliance-by-design architectures at scale. PO automation cuts processing costs 75-90% ($12.88→$2.36-$3.00 per invoice, per Ardent Partners and Engini benchmarks) and collapses cycle times from 17 days to 3-5 days, with 70-85% straight-through processing rates standard among mature deployments. Collections automation achieves similar maturity: AI-powered retry engines recover 35-65% of failed payments versus 15-25% for traditional dunning; voice AI agents achieve 76% portfolio resolution with 50% handling time reduction at $0.12/minute all-in (Leadlock); banking deployments achieve 48% cost-per-recovery reduction with 100% regulatory compliance. Agentic procurement negotiation—demonstrated at scale by Walmart's autonomous agent negotiating 2,000+ supplier contracts—extends automation beyond transaction processing to deal optimization. Oracle (Gartner Leader for Source-to-Pay 2025-2026) deployed 1,000+ agentic agents in Fusion Q2 2026, including collections workspace and sourcing command center; Sage Intacct shipped AI-driven 3-way matching GA globally (May 2026); CGI released three embedded collections capabilities; Esker delivers 90%+ touchless cash application; HighRadius serves 1,300+ enterprises with agentic collections. The defining tension remains execution: 80% of financial firms deploy AI (Cambridge Centre for Alternative Finance), yet procurement leaders show 58% expect material impact within 12 months while <50% are at scale—most remain at pilot/experimentation stage (Hackett Group/Zycus survey); 42% of enterprises scrapped AI initiatives amid total-cost-of-ownership overruns (200-400% above estimates). Regulatory fragmentation now governs adoption velocity more than technology capability: NYC SHIELD Rule (effective Sept 1, 2026) introduces 3-per-week contact cap and enhanced validation procedures for original creditors; FCC offshore call center rule (March 2026 vote) mandates onshoring, compliance reporting, and data-handling restrictions; FDCPA/TCPA enforcement accelerated with $69-$75M+ annual settlement volumes and class-action certification rates rising to 53%. Implementation friction and organizational capability—coupled with regulatory compliance complexity—remain the binding constraints, not technology risk.

CURRENT LANDSCAPE

Agentic AI infrastructure achieved production maturity Q2 2026 with vendor and enterprise validation across procurement and collections domains. Oracle (Gartner Leader for Source-to-Pay Suites) deployed 1,000+ agentic agents in Fusion Q2 2026, enabling autonomous collections management (Collectors Workspace), sourcing decisions, and compliance-aware exception handling across 54-country production base; AP AI Agent achieves 70-85% straight-through processing under $3 per invoice; Fusion 26C release (June 2026) added enterprise infrastructure—CLI development, debugger with breakpoints, policy models for deterministic compliance logic—advancing platform maturity for production-grade agentic scaling. Sage Intacct released AI-driven 3-way matching (invoice, PO, receipt) generally available globally (May 2026), with AI Gateway enabling external LLM integration—signals mid-market ERP maturity for AP automation. Agentic procurement negotiation demonstrated at Fortune 500 scale: Walmart deployed autonomous agents to negotiate 2,000+ supplier contracts achieving 3% cost savings, 35-day payment term extension, and 68-72% agreement success rates (April 2026). Collections vendors adopted agentic architecture: CGI Credit Studio (three AI capabilities: Call Summarization 30% after-call work reduction, Ask Cleo agent 20% productivity gain, Agent Assist 20% promise-to-pay improvement); Esker delivers 90%+ touchless cash application with customer outcomes (50% speedup, $2.3M deduction recovery); HighRadius (Magic Quadrant leader, 1,300+ customers, 30% productivity gain, 10% DSO reduction). Voice AI agents for collections reached 93% adoption among forward-thinking firms (up from 49% in 2023), with deployment metrics showing 76% portfolio resolution rates and 50% handling time reduction at $0.12/minute all-in (Leadlock); Zowie production case study confirms 60%+ inbound contacts resolved without human escalation and 3x improvement in after-hours payment arrangements. Cash application reached 99%+ accuracy with intelligent matching across partial/bundled payments; invoice dispute automation demonstrated across named customers (Danone $20M deduction recovery, Lucid.now $440K+ savings). Analyst validation: Hackett Group evaluated 118 S2P vendors across 500+ functional requirements (April 2026), establishing institutional vendor evaluation rigor; Cambridge Centre for Alternative Finance reports 80% of financial firms deployed AI (800+ respondent study, BIS/IMF/World Bank partnership).

Regulatory architecture now governs platform maturity and adoption velocity more than technology capability. NYC SHIELD Rule (finalized Feb 2026; effective Sept 1, 2026) expands debt collection oversight to original creditors, capping contacts at 3 per week across all channels (stricter than FDCPA's 7 federal), requiring any-channel dispute procedures and enhanced validation for medical debt. FCC's proposed offshore call center rule (March 2026 vote) mandates English-language proficiency testing, offshore volume caps (~30%), mandatory disclosure of offshore handling, data-handling restrictions for sensitive information, and FCC reporting obligations—driving vendor consolidation toward domestic AI solutions and compliance-embedded architectures. TCPA/FDCPA/FCRA enforcement accelerated: top 10 settlements totaled $69.1M (TCPA), $74.77M (FCRA/FDCPA) in 2025; class-action certification rates rose to 53% (up from 37% in 2024), increasing litigation risk and establishing predictability for vendor liability. Compliance-by-design enforcement—hard-constraint rule encoding vs. probabilistic guidance—became operational baseline. Agentic debt collection adoption faces multi-front regulatory pressure despite operational maturity: only 11% of third-party collection agencies deployed AI at scale despite 60% on active paths. Execution remains the binding constraint: Procurement AI readiness among 121 global professionals averages 2.1/5 (below deployment threshold per Suplari research); 58% of CPOs expect material agentic AI impact within 12 months but <50% are at scale (Hackett Group/Zycus); 83% lack enforced AI governance. Custom agentic agent deployments face realistic timelines of 12-18 months to production-safe maturity (stability across quarterly patches, observability infrastructure, consultant skill gaps); pre-built vendor agents (Collectors Workspace, Payables Agent) compress timelines to 4-8 weeks but remain constrained by organizational readiness. Organizations scrapped 42% of AI initiatives due to total-cost-of-ownership overruns (200-400% above estimates); 88% of AI projects fail pilot-to-production transition (IDC), highlighting organizational implementation barrier more acute than technology risk. Invoice-matching automation holds 90%+ accuracy in clean cases but sustains 85-90% vs 95%+ in pilots due to supplier data inconsistency at scale.

TIER HISTORY

ResearchJan-2019 → Jan-2019
Bleeding EdgeJan-2019 → Jan-2020
Leading EdgeJan-2020 → Apr-2025
Good PracticeApr-2025 → present

EVIDENCE (149)

— Oracle EVP Fusion Applications confirms 1,000+ agentic agents deployed in production across customer base; platform evolution from 'can we build an agent?' to 'how do we scale hundreds/thousands?' reflects industry maturity milestone.

— Zowie production deployment: 60%+ inbound contacts resolved without human escalation, 3x more payment arrangements outside business hours; McKinsey benchmark confirms digital-first collections see multi-percentage-point resolution gains.

— Oracle Fusion 26C platform maturity milestone: CLI tooling, debugger with breakpoints, policy models for deterministic compliance logic, long-running sessions, episodic memory—enterprise infrastructure for autonomous agentic collections/payments/PO automation.

— Analyst validation: Oracle's embedded agentic AI reduces integration burden (shared data, workflows, permissions, audit hierarchies across systems), enabling faster pilot-to-production transition; unmetered premium LLM access lowers adoption friction.

— AvidXchange Payment Automation for Workday launched on Workday Marketplace; enables payment automation and supplier workflows inside Workday Financial Management; signals ecosystem integration maturity for mid-market AP/payments automation.

— AvidXchange post-acquisition operational scaling: 99.2% invoice extraction accuracy, $21.5B Q2 2025 payment volume, 8,000+ customers, 1.5M+ suppliers; COO mandate targets agentic AI deployment scale.

— Critical practitioner assessment: Oracle agentic apps show impressive capability but production barriers remain real—stability across quarterly patches, observability immaturity, consultant skill gaps; realistic deployment timeline 12-18 months for custom agents.

— Collections AI adoption surged 49%→93% (2023-2025); Leadlock case studies show 76% portfolio resolution rates with 50% handling time reduction; cost USD $0.12/min vs. $36K+ human collector; addresses FDCPA compliance through hardcoded guardrails.

HISTORY

  • 2019: Cloud procurement platforms (Oracle Fusion, AvidXchange) demonstrated AI-powered approval agents and rule learning from historical transaction patterns; AvidXchange AvidPay network reached 220K+ active suppliers for electronic payments; EOS Group deployed ML-powered collections analytics platform across 26 countries; Tipalti secured $76M Series D funding for payables automation expansion.
  • 2020: RealPage integrated vendor payment automation via AvidXchange for real estate vertical; Tipalti scaled to $12B annual transaction volume (80% YoY growth) processing 4M+ suppliers; AvidXchange supplier network grew to 680K+ with $140B in annual transaction processing; independent analysts recognized AP automation vendors as mature market category with established customer bases.
  • 2021: Tipalti achieved $23B annualized transaction volume (83% YoY revenue growth) with new multi-subsidiary payables features for global FX management; AvidXchange expanded into construction vertical via TimberScan API integration; adoption remained constrained by organizational concerns about automation and collections staff displacement (43% executive worry reported by Deloitte).
  • 2022-H1: Tipalti surpassed $36B annualized transaction volume (100%+ YoY growth) with $270M Series F funding at $8.3B valuation; AvidXchange expanded ERP integrations (NetSuite, Sage 100 Contractor) to reduce manual PO-invoice matching; market consolidation continued around leading vendors while deployment customer friction (integration complexity, slow iteration) emerged as adoption constraint.
  • 2022-H2: Tipalti Card launched (September) with automated reconciliation; AvidXchange processed 42M invoices annually across 7,000+ customers with 60% cost and 70% time reductions; new integrations (QuickBooks Online, Oracle NetSuite) expanded mid-market reach; Symend raised $42M (November) for AI debt collection; three-way matching barriers (61% of late payments from admin errors) and implementation friction remained adoption constraints.
  • 2023-H1: Tipalti restructured with 11% workforce reduction signaling market pressure; AvidXchange expanded construction vertical with automated lien waiver management; Oracle Cloud Procurement gained government sector traction (UK local authority live Q3); regulatory scrutiny of AI debt collection increased; consumer preference for digital contact channels (75% avoid unknown calls) reshaped collections platform design.
  • 2023-H2: Enterprise deployments scaled: 50K-person mining/construction company transitioned to Oracle Cloud ERP with automated procurement; AvidXchange customer Remedy Medical Properties achieved 60%+ cost reduction and staffing avoidance; Oracle manufacturing case study demonstrated 15% cost savings. Regulatory compliance emerged as critical feature (Skit.ai voice AI for collections, state-level dunning laws); vendor stability concerns persisted (Tipalti platform issues reported by customers). Market consolidation continued around established vendors amid changing market dynamics.
  • 2024-Q1: Market adoption plateaued despite vendor growth: AvidXchange processed $19.9B annualized payment volume with 8,000+ customers, but product reliability issues and payment delays emerged in user reviews; collections AI adoption stalled (only 10% of US firms deployed, 60% considering) due to regulatory backlash (Robodebt scandal, CFPB AI chatbot warnings); Oracle ERP procurement deployments faced execution challenges (Birmingham City Council implementation required £5.3M remediation). Adoption barriers shifted from technical integration to organizational implementation capability and compliance governance.
  • 2024-Q3: Enterprise procurement deployments accelerated despite vendor headwinds: NHS Supply Chain went live with Oracle Fusion automating PO and payment processes; AvidXchange delivered 19.7M Q2 transactions (up 4.8%) across $20.6B payment volume (+10.4%), yet faced analyst revenue misses and FY24 guidance reductions, signaling valuation pressure. Collections automation remained cautious: systematic dunning demonstrated 54+ day faster collection cycles (60% DSO reduction), while compliance-focused AI voice platforms (Skit.ai) addressed CFPB concerns; market signaled need for human-AI hybrid approaches balancing automation with regulatory oversight.
  • 2024-Q4: Collections and payments automation demonstrated sustained operational scale: AvidXchange processed 20.2M Q3 transactions (+5.2% YoY) with $21.5B payment volume (+9.4%); HighRadius launched AI debt collection platform with 20%+ past-due reduction claims and 80% automation across 1,300+ enterprises. PO automation adoption accelerated (57% of procurement professionals working toward full automation) despite cautionary analyses on flawed workflow risks; regulatory compliance remains core differentiator as CFPB and Robodebt precedents shape platform design.
  • 2025-Q1: Enterprise procurement deployments scaled: Oracle Cloud Procurement expanded to 54-country EPCI energy sector deployments and building materials manufacturer integrations, while AvidXchange faced strategic evaluation (potential sale, analyst downgrades) despite Q4 2024 financial growth (19.9M transactions, 4.3% YoY, $21.9B volume). Collections AI adoption metrics clarified: TransUnion data showed 11% of third-party agencies using AI, 60% deploying, with primary use cases in payment prediction (58%), segmentation (56%), and behavior anticipation (46%). Critical signal emerged: generic AI platforms may harm collections outcomes vs. behavioral-science-informed approaches; 70% of adopting companies concerned about generic tool effectiveness. Vendor consolidation and platform specialization accelerate as market matures.
  • 2025-Q2: Strategic consolidation reshaped vendor landscape: AvidXchange acquired by TPG and Corpay for $2.2B (May), raising post-acquisition innovation concerns despite sustained transaction volumes; Oracle named Gartner Leader for Source-to-Pay Suites (April), signaling analyst confidence in embedded AI capabilities. Collections deployments delivered quantified ROI—McKinsey data: 40% ops reduction, 10% recovery gains, 30% satisfaction improvements; PO automation case studies showed sub-minute AI-enabled reconciliation (Ghost warehouse), while broader adoption metrics confirmed 57% of procurement professionals working toward full automation. Procurement software benchmarks: 6:1 ROI, 50% sourcing time reduction, 15% cost savings, 30% dispute reduction. AR automation deployment timelines stabilized at 4-12 weeks basic, 12-20 weeks end-to-end, though Gartner survey found 61% of finance functions lacked AI implementation plans—signaling adoption constraints despite documented efficacy.
  • 2025-Q3: PO automation acceleration continued with real-world deployments demonstrating concrete value: Canadian distributor successfully deployed Power Automate-based PO processing with custom AI extraction and human-in-the-loop approval workflow (September), while 80% of global CPOs reported plans to deploy generative AI for procurement within 3 years, with GenAI capable of automating 80% of repetitive PO/invoice tasks. AvidXchange maintained transaction volume momentum (20.1M Q2, +1.8% YoY) across $21.5B payment processing despite pending TPG/Corpay acquisition integration uncertainty. Collections automation adoption faced intensifying regulatory scrutiny: CFPB, GDPR, and international data protection regimes (DPDPA, UAE localization) created material compliance barriers for cross-border AI collections deployments; 2024 $12M enforcement penalty for unlawful credit reporting signaled escalating enforcement risk. Manual PO matching remained process bottleneck (~11 hours per order, frequent early payment discount misses), creating continued demand for three-way/four-way matching automation. Overall, Q3 2025 demonstrated PO automation momentum offset by collections automation compliance tightening and post-acquisition uncertainty in leading vendor landscape.
  • 2025-Q4: Real-world deployments validated ROI at scale: Nucleus Research case study documented food/agriculture manufacturer achieving $24M long-term savings from improved procurement automation; distributor reduced financial close from 5 to 4 days. AvidXchange customer deployments confirmed production impact (Piedmont Service Group saved $30K paper costs + $20K early-payment discounts; Fox Theatre reduced processing time to <48 hours, saving $80K annually). Strategic shift emerged in market: AI debt collection adoption transitioning from cost-cutting to compliance-first focus, with U.S. banks and fintechs prioritizing regulatory control and audit trail capabilities. Procurement investment acceleration confirmed: 72% of finance departments currently investing in AI/ML with 53% planning increased 2026 spend; 80% of global CPOs planning generative AI deployment within 3 years targeting 80% automation of repetitive tasks. Critical adoption gap persisted: 80% of finance executives rated AR automation strategic priority yet only 3% achieved full automation; 11% of debt collection agencies deployed AI despite 60% on active paths. Vendor landscape uncertainty continued: AvidXchange post-acquisition integration costs ($6.4M disclosed) and innovation roadmap questions raised risk concerns; compliance-focused platforms (Skit.ai, Symend, Collect.ai) expanded amid regulatory tightening. Implementation barriers remained: AR automation required 4-12 weeks basic to 12-20+ weeks full deployment; budget constraints and IT backlog persisted as primary adoption obstacles.
  • 2026-Jan: Early 2026 showed continued operational scaling alongside significant vendor credibility challenges: Fortune 500 manufacturer achieved 70% PO automation via unattended workflows combining email/ERP extraction (Altimetrik case study); collections vendors refined competitive positioning as AI retry engines demonstrated 2-4x higher recovery rates vs. dunning emails. However, leading vendor AvidXchange faced mounting operational friction: G2 reviews documented slow support and payment delays; BBB complaints from January 2026 reported 3-month payment gaps despite platform showing processed status, signaling adoption risks from operational failures. Landscape reflected dual tension: core automation technology maturity (successful PO processing deployments, proven recovery improvements) vs. vendor platform instability limiting customer confidence. Market consolidation continued with TPG/Corpay integration timeline uncertain; compliance-first collections platforms maintained differentiation through regulatory transparency as federal enforcement priorities intensified.
  • 2026-Feb: February confirmed divergence between adoption intent and execution capability: Oracle (2026 Gartner Leader for Source-to-Pay) maintained analyst confidence with embedded AI agents for PO automation; 94% of procurement executives using GenAI weekly yet only 4% achieved large-scale deployment (adoption-execution gap). AvidXchange released integration updates with NetSuite, Dynamics GP, Sage Intacct signaling product viability despite TPG/Corpay integration uncertainty. Critical shift in market narrative: major analyses revealed structural ROI challenges—95% of enterprise GenAI pilots failed to deliver returns in 6 months; 42% of companies scrapped AI initiatives (double prior year); total AI ownership costs run 200-400% above initial quotes. Collections-side analysis showed AI adoption stalling in effectiveness: 78% of CFOs deployed AI for billing anomalies but only 53% saw modest cash conversion gains. Ardent Partners benchmarked PO automation ROI at 78% cost reduction ($12.88→$2.78 per invoice) and cycle time collapse (17 to 3 days), confirming technology maturity while highlighting implementations struggling with change management and cost overruns.
  • 2026-March: March 2026 data confirmed sustained platform maturity with emerging agentic capabilities: Oracle announced embedded AI agents for Fusion suite automating PO/invoice data entry and execution tasks (Reuters, March 24); independent analysis (Monk, March 11) benchmarked 42 AR platforms across 200+ finance teams showing 20-40% DSO reduction and 50% cost savings with 3-6 month payback. Named customer deployments demonstrated continued production scale: Red Bull (43% productivity gain, $6.2M working capital recovery), Konica Minolta (9-day DSO reduction, $3.5M savings), Danone (96% forecast accuracy, $20M deduction recovery); peer-reviewed case study of Chinese SME (DK Company) showed 25.7% DSO reduction and RMB 1.5M cash recovery. Emerging signal: Billtrust (March 24) released agentic AI features (Agentic Email, Agentic Procedures, Agentic Credit Lines) with 50% DSO improvement claims, indicating vendor response to LLM/agentic market momentum. Critical constraint persists: Big 4 consulting guidance (Forvis Mazars, March 23) flagged over-automation risks in AP, warning that naive 3-way matching without upstream process design creates bottlenecks rather than enabling automation. Market maturity confirmed but organizational implementation capability remains limiting factor for faster adoption.
  • 2026-Apr: Agentic AI maturity accelerated with major vendor capabilities and analyst validation: Oracle (April 9) launched 12 Fusion Agentic Applications including Collectors Workspace for autonomous collections management and Sourcing Command Center for procurement decision-making. AP AI Agent documented at 70-85% straight-through processing with cost under $3/invoice; manufacturing customer case study showed exception-handling reduction from 60% to under 20% of AP team time in six months. CGI Credit Studio released three embedded AI capabilities for collections (April 2026): Call Summarization (30% after-call work reduction), Ask Cleo agent (20% productivity gain), and Agent Assist (20% promise-to-pay conversion improvement). Ramp published seven named AP automation case studies (April 2026) showing 80%+ invoice processing time reductions across healthcare (Hospital Association of Oregon: 10-hour to minute-level processing), arts (The Second City: 2x faster throughput), and logistics (REVA Air Ambulance: 15-20 min to under 3 min per invoice). Analyst validation expanded: Forrester published a report on agentic AI maturity across AP functions; Billtrust released research documenting 25 agentic AR use cases with adoption metrics from customer deployments; ML-powered collections platforms demonstrated real-world automation of payment delay prediction and risk scoring (Lucid.now). Practitioner analysis cut through vendor positioning, distinguishing genuine AP automation capability gains from marketing hype while confirming cost-economics (sub-$3 per invoice) remain solid. Collections case studies demonstrated regulatory-safe deployment: banking sector achieved 48% cost-per-recovery reduction with 100% FDCPA compliance (iTuring.ai case). Procurement adoption intent remained at 90% of leaders implementing or planning AI agents, yet the execution gap persisted: practitioners noted critical distinction between capability maturity (vendors delivering) and organizational readiness (only 4% achieving large-scale production), with cost-of-ownership and implementation friction constraining faster rollout.
  • 2026-May: Esker reported 90%+ touchless cash application rates with customer-verified 50% processing speedup and $2.3M deduction recovery, while AI agents handling invoice disputes demonstrated production outcomes at scale (Danone $20M deduction recovery, Lucid.now $440K+ savings, LedgerUp 95% auto-match). Cambridge Centre for Alternative Finance's 2026 report (800+ respondents, BIS/IMF/World Bank partnership) confirmed 80% of financial firms have deployed AI broadly, providing macro validation for the sector's automation trajectory. Oracle's 1,000+ Fusion agentic agents in production (70% customer base) and Walmart's autonomous negotiation agents across 2,000+ suppliers (3% cost savings, 35-day payment term extension) marked the period's headline deployments; Hackett Group evaluated 118 S2P vendors across 500+ functional requirements, providing institutional vendor differentiation at scale. AvidXchange released new AI agents — Approval Agent (historical pattern-based routing) and PO Matching Agent (line-item automation) — with 99.2% invoice extraction accuracy and integrated fraud detection, extending the platform's automation depth despite ongoing post-acquisition integration uncertainty; Oracle Fusion's 26B release similarly added a dedicated Payments Agent alongside the Payables Agent, confirming major ERP platforms are now shipping agentic capability across the full AP lifecycle rather than single-workflow automation.
  • 2026-Jun: Oracle announced 600+ pre-built Fusion agentic agents GA including AP invoice matching, AR collections follow-up, and PO management agents; Sage Intacct released AI-driven 3-way matching globally, signaling mid-market ERP maturity for AP automation. Collections voice AI adoption confirmed at 93% among forward-thinking firms (up from 49% in 2023) with Leadlock documenting 76% portfolio resolution rates and 50% handling time reduction at $0.12/minute all-in versus $36K+ for human collectors; Ardent Partners benchmarks show 77% per-invoice cost reduction ($10.18→$2.36) through semantic extraction and AI-driven matching. Regulatory pressure intensified: NYC SHIELD Rule (effective September 1, 2026) caps contacts at 3/week across all channels and imposes stricter validation—stricter than FDCPA—while Hackett Group survey (250+ CPOs) found 58% expect material agentic AI impact within 12 months but under 50% are at pilot stage, and Suplari primary research (121 procurement professionals) measured AI readiness at 2.1/5 with 83% lacking enforced governance. Oracle Fusion 26C platform infrastructure reached a maturity milestone with CLI tooling, a debugger with breakpoints, and policy models for deterministic compliance logic—enterprise infrastructure enabling production-grade agentic scaling for collections and PO workflows; AvidXchange Payment Automation launched on Workday Marketplace, extending mid-market AP automation to Workday Financial Management customers. Practitioner assessment noted a realistic 12-18 month timeline to production-safe maturity for custom agentic agents despite impressive capability, with observability immaturity and quarterly patch stability flagged as the binding operational constraints.

TOOLS