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

Recruitment content & compensation modelling

GOOD PRACTICE

TRAJECTORY

Stalled

AI that generates job descriptions, interview questions, offer letters, and models compensation packages for hiring workflows. Includes bias-checked JD generation and market-rate compensation analysis; distinct from candidate sourcing which finds candidates rather than creating recruitment materials.

OVERVIEW

AI-generated recruitment content and compensation modelling has reached good-practice maturity with mainstream adoption in content generation and measured but selective adoption in compensation modelling. Job description generation with bias detection is now standard enterprise practice, while compensation analytics platforms operate at massive scale—beqom alone manages 5M+ employees across 40+ S&P 500 companies. The practice is defined by a widening gap between vendor capability (agentic compensation intelligence, real-time pay equity analytics, multi-scenario modelling) and organizational deployment discipline (human-centered decision-support only, never autonomous). Regulatory drivers are closing this gap: state-level laws (Illinois, Colorado, New York) and the EU Pay Transparency Directive now mandate bias audits, employee notice, and disparate impact accountability—converting compliance from optional to mandatory.

The practice covers two distinct capabilities sharing workflow logic: content generation (job descriptions, interview questions, offer letters, requisitions) and pay modelling (market-rate benchmarking, equity analysis, scenario planning, offer guidance). Content generation has commoditized—66% of HR professionals deploy AI for JDs, 87% of companies embed AI into hiring, Grammarly ships free generation to 150M+ users. Compensation modelling remains bifurcated: vendors ship sophisticated agentic agents with ML-based recommendations and bias elimination; organizations adopt conservatively (only 2% actively use AI for pay decisions, 50% piloting) with human approval required for all compensation outcomes. The tension between capability and confidence is not vanishing but shifting: as regulation raises the bar for bias auditing, documentation, and impact assessment, the organizational case for human oversight strengthens—not because AI is weak, but because the legal and reputational cost of discriminatory impact (regardless of intent) is now material. Emerging operational failures expose real deployment gaps: compensation modelling tools assume uniform role taxonomies but AI skill-specific roles command 50-100% equity premiums vs. benchmarked roles, breaking budget models mid-planning; AI recruitment content at scale requires human review to avoid embedding discriminatory logic; and fraud signals emerging in recruitment (41% of candidates exploit prompt injection, 46% report decreased trust) demand stronger control architectures.

CURRENT LANDSCAPE

Recruitment content generation has achieved mainstream saturation, yet adoption masks a persistent outcomes gap. Job description generation penetration: 66% of US HR professionals deploy AI for JDs (up from 27% in 2022, representing sustained growth); 87% of companies incorporate AI into hiring workflows; Grammarly (150M+ users) ships free no-signup JD generation signaling full commoditization. However, 88% of HR leaders report NOT realizing significant business value from their AI investments—indicating that rapid adoption has outpaced implementation rigor and value realization. JD content is evolving structurally: PwC analysis of 1B+ job postings shows AI-exposed entry-level roles are 7x more likely than non-exposed roles to require traditionally senior skills (seniorisation), with AI specialist roles growing 68.9% and commanding 62% average wage premiums—evidence that market-driven skill demands are reshaping recruitment content generation. Platform consolidation deepens: SAP SuccessFactors announced GA Joule Assistants (May 2026) for recruitment workflow orchestration (matching through interview coordination) and payroll automation; Workday and Compa certified their AI Analyst Agent integration (May 2026) embedding compensation modelling inside production HCM workflows. Interview question generation is now embedded in ATS platforms with documented 25-30% bias reduction and 40-50% improved relevance, and AI-assisted gender-neutral JD rewrites increase application rates among underrepresented groups. Field deployment impact: TalentBridge Solutions (2,000-person consulting firm) reduced screening time per role from 23 to 6 hours (74% reduction) with 11 percentage point retention gain, validating AI automation of high-volume recruiting tasks; recruiters using AI-assisted tools are 9% more likely to make quality hires; AI skill premiums now documented at 27% salary increase for AI-fluent workers; B2B sales teams deployed AI-driven pay-for-performance (71% adoption rate, June 2026) with measured outcomes: quota attainment improved 41% to 58%, turnover dropped 18% to 8%. Emerging friction points: candidate-side fraud (41% of candidates admit prompt injection to bypass AI screening) and eroding trust (46% of job seekers report decreased confidence in hiring; only 31% of CHROs report strong fraud controls) signal that recruitment content generation at scale requires robust human oversight and validation infrastructure.

Compensation modelling remains sharply bifurcated despite agentic capability advancement. Vendor innovation accelerates: beqom's Pay Intelligence and new entrants (Trusaic R.O.S.A., Evenpay ML integration) enable multi-scenario modelling, real-time pay equity detection, and cost-effective remediation guidance. Yet organizational adoption remains selective and governance-centric: Korn Ferry survey (4,000+ companies, March 2026) shows 50% developing or piloting AI pay processes, but only 2% actively use AI for pay decisions; 56% not considering it; 84% of organizations still rely on general-purpose ChatGPT rather than compensation-specific tools; 60% of compensation professionals remain skeptical about automation. The critical barrier is not capability but implementation discipline: Pave survey (525+ compensation leaders, June 2026) documents the "say-do gap"—80% of companies with documented compensation philosophy are NOT using AI for pay recommendations; 75% with integrated data are NOT using AI for pay-equity analysis—revealing that data readiness and governance, not tool immaturity, constrain deployment. Governance is hardening as the enabling layer: 93% of compensation leaders now involve C-suite, IT, and finance in compensation software decisions (historically HR-owned), signaling institutional emphasis on oversight and integration discipline. This reflects a structural shift: compensation teams are moving from annual survey-based benchmarking to real-time AI analytics (68% of postings now include salary ranges, up from 45% in 2023), and organizations are restructuring compensation around AI-driven models—B2B sales teams report 71% adoption of AI-driven pay-for-performance with specific design shifts (OTE splits 53/47 base/variable, quota +30–55%, AI fluency premiums 4–5%) and measurable outcomes (22–31% higher attainment, 17–24% lower attrition, 41% faster sales cycles). Critical operational failures are emerging: compensation modelling tools designed for standard benchmarks break when organizations deploy AI-specific roles that command 50-100% equity premiums over baseline roles, forcing role-by-role recalibration mid-planning cycle (evidence from Equity People). Vendor guidance (beqom, May 2026) explicitly recommends against end-to-end autonomous agents, advocating instead for narrow-scope, task-specialized agents with human formula control as the integration point—signaling that compensation AI maturity depends not on capability breadth but on governance architecture. Market scale remains robust: beqom manages 5M+ employees across 40+ S&P 500 companies (Total Energies, Allianz 100k employees, Lowe's), with Lowe's platform delivering processing time reduction from 12+ hours to <2 hours for complex variable compensation; EU Pay Transparency Directive compliance workflows (June 1, 2026 deadline) are now using frontier LLMs to draft mandatory pay gap reports at scale for 250+ employee organizations. Organizational readiness analysis (manufacturing sector study, May 2026) reveals that AI integration alone is insufficient: companies with high technical AI deployment but low organizational transparency cultures see persistent disclosure gaps, indicating that maturity requires institutional change (governance, audit, transparency culture) alongside technology adoption. EU AI Act compliance now entering enforcement phase (August 2026): German firms using AI for salary recommendation and compensation decisions face mandatory transparency audits and risk-assessment requirements, signaling that regulatory compliance will drive investment in governance-first compensation modelling infrastructure.

Regulatory framework has shifted from guidance to binding obligation and is the primary adoption driver. State-level: Illinois HB 3773 (effective Jan 1, 2026) bans AI with discriminatory effects regardless of intent, mandates notice to employees when AI affects employment decisions, and imposes $5,000 per violation penalties. Four states enacted conflicting employment AI laws (May 2026 analysis); federal EEOC guidance was removed leaving only state-level standards. EU context: EU Pay Transparency Directive (effective June 2026) mandates pay gap reporting for 250+ employee organizations; EU AI Act (effective August 2026) applies to employment decisions with strict bias testing and auditability requirements. Enforcement is live: DOJ settled its eighth case (March 2026, reaffirmed May) against an IT firm for AI-generated job ads illegally excluding US citizens—demonstrating that content generation at scale requires human review and compliance infrastructure to avoid discriminatory outcomes. Vendors have responded by embedding bias audits and compliance controls (disparate impact testing, feature attribution analysis, 80/20 rule detection) as core platform features. The critical adoption barrier is no longer capability but governance: organizations must operationalize bias auditing, impact assessment, and human approval workflows to deploy legally defensible AI in recruitment and compensation.

TIER HISTORY

ResearchJan-2023 → Jan-2023
Bleeding EdgeJan-2023 → Jan-2025
Leading EdgeJan-2025 → Apr-2025
Good PracticeApr-2025 → present

EVIDENCE (115)

— 71% of B2B sales teams deployed AI-driven pay-for-performance compensation with structural model shifts: quota baselines +30–55%, OTE splits 53/47 base/variable, AI fluency premiums 4–5%, documented 54% cost-per-opportunity reduction via AI.

— Kory White analysis of 2027 B2B comp restructuring away from activity KPIs (eliminated by 68% of firms per Gartner) toward outcomes (net-new logos, multi-threaded deals); proposes three-bucket model with buying-committee multiplier and AI Override Clause.

— PulseRevOps CS compensation design uses AI-informed metric construction (75/25 base/variable, NRR 50%/GRR 30%/health 20%); clawback governance embedded; demonstrates outcome-based compensation modelling at scale.

— PulseRevOps compensation design guidance for AI-augmented AE roles shows 22–31% higher attainment, 17–24% lower attrition, with specific model shifts: base 55–65% OTE, variable 35–45%, quota +30–55%, AI-fluency component 3–7% variable.

— beqom survey of 178 US compensation leaders: 93% involve C-suite/IT/finance in comp software decisions; 52% require data privacy safeguards before adopting AI—governance-first adoption pattern is institutional norm.

— PwC analysis of 1B+ job postings shows AI-exposed entry-level roles 7x more likely to require senior skills (seniorised JD content); AI specialist roles 68.9% growth; roles requiring AI skills carry 62% average wage premium—evidence of market-driven recruitment content redesign.

— Pave 525+ compensation leader survey reveals critical adoption gap: 80% with documented philosophy not using AI for recommendations; 75% with integrated data not using AI for equity analysis—data readiness and governance, not capability, is the barrier.

— German firms using AI for salary recommendation and compensation decisions face August 2026 compliance deadline (EU AI Act); active ecosystem (EY audit agents, Gemini integration, salary negotiation chatbots) shows live deployment with regulatory compliance activity.

HISTORY

  • 2023-H1: Early adoption of generative AI (ChatGPT) for job descriptions and interview questions (49% of business leaders); AI compensation tools (Payscale Verse, Compensation IQ) achieving market recognition. Academic research surfaces significant gender bias in salary negotiation AI and inconsistency in pay prediction models, underscoring fairness risks.
  • 2023-H2: Ecosystem expansion with new JD generators (Intrvuz, 9am) in beta/GA; Payscale reports 235% ROI from compensation management software. Vendor consolidation accelerates: beqom acquires PayAnalytics to strengthen pay equity offering. Industry focus remains on bias mitigation in recruitment content, with Textio emphasizing fairness-first approach.
  • 2024-Q1: Adoption progress on JD generation (21% of organizations developing or using AI); beqom reports efficiency gains (95% spreadsheet reduction, 25% retention increase). Regulatory landscape tightens: UK government publishes responsible AI guidance for recruitment; Ontario and California consider AI disclosure laws. Industry sentiment mixed: 49% of HR leaders optimistic about AI compensation, 17% pessimistic due to bias and sophistication concerns. Deployment barriers clarify: experts caution AI insufficient for autonomous pay decisions.
  • 2024-Q2: Leading vendors ship measurable product enhancements: Payscale releases AI-powered job summaries and compensation planning features (2,200+ active summaries); beqom publishes case studies showing real deployments (LähiTapiola 50% efficiency gain, Breitling fair pay leadership). Mercer and Compa articulate vision for personalized compensation and skill-based pay modelling. Academic research (May 2024) documents persistent fairness and measurement gaps in AI-driven recruitment and compensation, reinforcing need for human oversight and audit controls.
  • 2024-Q3: Adoption momentum accelerates: ISG and Gartner confirm AI as standard in compensation planning and recruitment (76% of companies planning implementation); Randstad reports 81% of HR leaders using or exploring GenAI recruiting. Job description generation now mainstream use case; Mercer quantifies 52% of rewards team workload as AI-automatable. Critical limitations surface: content quality analysis warns GenAI lacks production-phase precision and consistency; practitioners continue treating AI as drafting aid requiring substantial human review rather than autonomous tool.
  • 2024-Q4: Adoption reaches late mainstream with high absolute numbers: 91% of compensation leaders using AI in workflows, 65% of HR professionals using AI for job description generation, 3.5x growth in GenAI job posting mentions (Indeed). Public sector validation: UK Ministry of Defence deploys Textio for advert optimization (December). Vendor ecosystem remains active but shows signs of scaling strain: major players ship product updates while facing sustainability challenges. Deployment barriers intensify: industry analysis warns of bias risks (30% of AI models), compliance complexity (wage/hour law, dynamic pricing), and sustained human review requirements. Broad workforce adoption steady but limited: 24% of employees use GenAI weekly, with 1-8% of work hours AI-assisted, indicating adoption breadth without transformative depth.
  • 2025-Q1: Adoption consolidates into mature mainstream with vendor specialization: beqom launches AI Pay Prediction (multivariate models for salary recommendations and retention risk) and BYOM (bring-your-own-model) capabilities, signaling market demand for customizable compensation modelling. Payscale survey (3,595 respondents) shows 47% using AI for JD generation, 48% optimistic on pay equity monitoring, 45% on compliance enforcement—organizational focus shifts to fairness and regulatory alignment. ISG's 2025 analyst assessment lists 10 major vendors and 11 emerging competitors, confirming ecosystem maturity. However, organizations remain cautious: 52% explicitly plan not to replace humans with AI, indicating sustained reliance on human oversight. Vendor sustainability concerns persist (Textio layoffs), and skills-based compensation modelling remains pilot-stage. Core tension deepens: vendors ship sophisticated AI capabilities while organizations deploy as decision-support tools requiring substantial human judgment.
  • 2025-Q2: Capability expansion with agentic AI and advanced analytics: beqom releases Pay Intelligence (May) with ML-driven recommendations and bias elimination; Compa launches AI Agents (June) for automated market intelligence and offer guidance. Real-world deployments documented: beqom case studies confirm adoption at Total Energies, Allianz (100k employees), and Lowe's for pay equity and performance-linked compensation. Market consolidation accelerates around enterprise platforms; specialized AI compensation tools now manage compensation for millions of employees globally. Recruitment content generation remains mainstream with expanded vendor ecosystem. Organizational deployment conservative: AI positioned as augmentation tool requiring human judgment rather than autonomous decision-maker. Skills-based and agentic automation remain promising but implementation-limited.
  • 2025-Q4: Vendor innovation continues through year-end: Payscale launches Smart Price and Compass (Nov) for advanced job pricing and AI-driven compensation intelligence. ISG's 2025 TCM Buyers Guide (Nov) evaluates 23 providers across four platform categories, confirming sustained vendor ecosystem maturity. Industry bodies (SBAM) document continued benefits in job description generation and compensation analysis, while cautioning against algorithmic bias and implementation risks. Organization adoption patterns hold: AI remains positioned as decision-support augmentation tool with human review required. Job description generation consolidated as best-practice standard; compensation modelling capabilities mature but deployment conservative due to bias, compliance, and change management concerns.
  • 2026-Jan: Deployment evidence and organizational adoption sentiment strengthen in early 2026. Real-world case study (Cheyenne Regional Medical Center) confirms Payscale Paycycle delivering $80K+ savings and 95% high-performer retention, validating production-stage compensation planning deployment. Organizational sentiment remains cautiously optimistic: 71% of HR professionals want AI for compensation benchmarking, 68% trust AI for pay recommendations, but only 8% of CHROs believe managers have AI skills—signaling sustained capability gaps despite strong optimism. Compensation strategy shifts toward AI-driven precision targeting: WTW forecasts 2026 merit budgets consolidating at 2-3%, with organizations shifting from blanket increases to AI-enabled targeted investments tied to skills and performance. Regulatory pressures intensify significantly: California (AB 2930), Colorado (SB 24-205), Illinois (HB 3773), and EU AI Act (effective Aug 2026) mandate bias testing, transparency, and human oversight for AI employment systems, shifting equity from organizational aspiration to binding compliance standard. Deployment barriers persist: critical assessments warn of risks from bias amplification, transparency gaps, and data quality issues requiring sustained human judgment and oversight. Market scale remains robust: beqom's platform manages compensation for 5M+ employees across 40+ S&P 500 companies, demonstrating mainstream enterprise adoption of AI-enabled compensation infrastructure.
  • 2026-Feb: Adoption metrics and product innovation accelerate amid implementation caution and regulatory tightening. Payscale's 2026 survey (3,413 respondents) shows 61% of organizations updated roles to include AI skills, but compensation lags: only 14% offer higher base pay for AI skills, 9% offer long-term incentives—revealing widespread AI skills adoption without proportional pay elevation. Compensation AI adoption remains cautious: only 19% of HR professionals use AI for market pricing/benchmarking, 12% for compensation philosophy, with 28% cautious/hesitant about any AI compensation decisions, indicating adoption gains in job descriptions offset by trust gaps in pay decision automation. Product innovation continues: new entrants (CompSense.ai in India) and refined tools (Cangrade's JD Decoder) expand ecosystem; vendor focus shifts to enterprise compliance and governance—ZBrain and peers emphasize standardisation, bias detection, and HRIS integration for production-grade JD automation. Regional adoption patterns: beqom's European analysis shows 38% of European companies outgrowing legacy compensation systems, with 35% leveraging EU Pay Transparency Directive for modernisation, signaling infrastructure-driven adoption alongside regulatory drivers. The fundamental tension persists: vendors ship AI capability innovations while organizations deploy selectively with sustained human oversight, particularly in compensation decision-making where regulatory compliance requirements (EU AI Act, state-level testing mandates) increasingly constrain algorithmic autonomy.
  • 2026-Mar: Broad organizational adoption accelerates across both content and compensation domains. Korn Ferry survey (4,000+ companies in 133 countries) shows 50% of organizations developing or piloting AI processes for pay decisions, yet 64% of compensation professionals remain unsure what premium to offer—indicating widespread deployment without operational confidence. Job description generation reaches 62% penetration among US talent professionals (up from 27% in 2022); 85% of employers now adopt skills-based hiring. Employee sentiment strengthens: 67% of job candidates prefer companies using AI for pay decisions, though 33% trust AI over managers and hybrid human-AI models prove most trusted. Product innovation continues agentic direction: Trusaic's R.O.S.A. remediation agent models hundreds of pay scenarios for gap closure; Evenpay's ML-based platform provides real-time salary recommendations integrated with HRIS systems. Critical limitation surfaces: DOJ settlement (eighth under Protecting US Workers Initiative) with IT firm for AI-generated job ads discriminating against US citizens — the clearest signal yet that content generation at scale requires human review to avoid embedding discriminatory logic. Regulatory and market infrastructure drivers intensify: 11 US states now require salary transparency; EU Pay Transparency Directive (effective 2026) accelerates modernization of compensation systems. The practice's maturity is evident in scale—62% of hiring professionals deploying AI tools—yet deployment discipline concentrates in decision-support and compliance rather than autonomous decision-making.
  • 2026-Apr: Job description generation has reached mainstream saturation — 66% of US HR professionals now use AI for JDs (up from 27% in 2022), 87% of companies incorporate AI into hiring workflows, and Grammarly (150M+ users) launched a free no-sign-up JD generator signalling full commoditisation. Compensation AI adoption remains sharply bifurcated: while Decusoft shipped GA agentic compensation intelligence (pay equity analysis, scenario modelling) and 10 major platforms now bundle AI compensation agents, only 2% of organisations actively use AI for pay decisions — with 56% not considering it — confirming that content generation is good practice but compensation automation remains decision-support territory requiring human authority.
  • 2026-May: Enterprise ecosystem consolidation and organizational adoption of compensation modelling accelerates. Workday and Compa announced certified integration enabling AI Analyst Agents inside production HCM workflows—signaling major platforms embedding specialized compensation modelling as core infrastructure. Organizations are actively restructuring compensation around AI skills, with documented 27% salary premiums for AI-fluent workers, and real-world implementation of AI-driven pay-for-performance in B2B sales showing 71% adoption with measurable outcomes (quota attainment 41%→58%, turnover 18%→8%). Market intelligence infrastructure shifts structurally: compensation teams moving from annual survey-based benchmarking to real-time AI analytics (68% of job postings now include salary ranges, up from 45% in 2023). Industry governance frameworks mature: beqom and WTW co-published "Intentional AI" framework formalizing safe, deterministic, auditable AI in compensation decisions. However, critical adoption barriers persist: 50%+ of leadership pressure organizations to adopt compensation AI, yet only 16% use compensation-specific tools (84% still rely on general ChatGPT), and 60% of compensation professionals remain skeptical about full automation—confirming deployment concentrates in decision-support augmentation, not autonomous decision-making.
  • 2026-Jun: Vendor consensus hardens against autonomous compensation agents: beqom guidance explicitly recommends narrow-scope, task-specialized agents with human formula control as the integration point, and documented operational failure emerged as AI role premiums of 50–100% over baseline benchmarks broke headcount models mid-planning cycle — signaling compensation tools require role-granular data governance. A Pave survey of 525+ compensation leaders reveals a critical say-do gap: 80% with a documented compensation philosophy are not using AI for pay recommendations, and 75% with integrated data are not using AI for equity analysis — confirming that data readiness and governance, not tool maturity, remain the primary constraint on compensation AI adoption. Large-scale enterprise deployment continues: Lowe's (300k employees, 2,200+ stores) deployed beqom platform reducing commission-processing time from 12+ hours to under 2 hours with full audit trails, while a beqom survey of 178 US compensation leaders found 93% now involve C-suite, IT, and finance jointly in compensation software decisions — a historically HR-owned function — signaling governance-first adoption is now the institutional norm. B2B sales teams show the most mature deployment: 71% have adopted AI-driven pay-for-performance with structural shifts (OTE splits 53/47 base/variable, quota baselines up 30–55%, AI fluency premiums 4–5%) and measured outcomes including 54% cost-per-opportunity reduction. TalentBridge Solutions case study confirms recruitment content generation ROI (74% screening time reduction, 11pp retention gain), while candidate-side AI fraud escalated sharply: 41% of candidates admit using prompt injection to bypass AI screening, with only 31% of CHROs reporting strong fraud controls. Regulatory pressure intensifies: German employers using AI for salary recommendations face an August 2026 EU AI Act compliance deadline for transparent pay algorithms. SHRM 2025 Talent Trends (1,000+ respondents) confirms 66% of HR professionals use AI for writing job descriptions, 89% report time savings, but the outcomes gap persists with 88% of HR leaders reporting they have not realized significant business value — reinforcing that adoption breadth has outpaced implementation discipline.

TOOLS