Generative AI in Key Business Functions

8.1 The GenAI-Powered Value Chain

Before we get to individual functions, it helps to see the whole picture. Borrowing Michael Porter's classic value chain framework, we can trace how AI works its way into both primary activities (those directly involved in creating and delivering products and services) and support activities (those that enable the primary functions).

Figure 8.1: The GenAI-Powered Value Chain

The framework shows AI running horizontally across the business, from raw material sourcing to customer service, while also supporting functions like HR, R&D, and accounting. Seeing the full spread matters. A company that treats AI as a handful of isolated point solutions will build a very different, and much weaker, strategy than one that plans for the whole chain.

Primary Activities: AI Across the Operational Core

1. Inbound Logistics: Intelligent Supply Chain Management. Generative AI is reshaping how organisations manage incoming materials and information. In demand forecasting, a global electronics manufacturer can analyse historical sales data, social-media trends, and economic indicators to sharpen component-demand forecasts and cut inventory carrying costs; the size of the gain depends entirely on the baseline, and any vendor quoting a universal number is quoting marketing. Supplier communication becomes effortless across borders: AI automatically generates and translates purchase orders, quality specifications, and compliance documentation across more than 40 languages. Quality inspection benefits from this approach. Computer-vision models powered by generative AI create synthetic defect images to train inspection systems, allowing manufacturing flaws to be caught in real time at receiving docks. And route optimisation produces optimal shipping routes that account for weather, traffic, fuel costs, and delivery deadlines, surfacing detailed scheduling reports for warehouse managers.

2. Operations: AI-Enhanced Production and Service Delivery. The operational core is where some of the biggest AI gains show up. Process optimisation turns simulation into a strategic tool, a chemical plant can use AI to generate thousands of virtual experiments testing temperature, pressure, and catalyst combinations and narrow the physical experiments to the handful worth running. Predictive maintenance derives schedules directly from sensor-data patterns, and when anomalies appear the AI creates detailed maintenance reports with visual diagrams showing likely failure points. Quality-control documentation shrinks from days to minutes as AI generates comprehensive quality reports, root-cause analyses, and corrective-action recommendations from production-line data. When processes change, AI rolls out updated standard operating procedures in multiple formats, text, video, interactive guides, tailored to different worker skill levels and languages. And in production planning, AI creates optimised schedules that account for machine capacity, workforce availability, material lead times, and customer deadlines, surfacing them as detailed Gantt charts and resource-allocation plans.

3. Outbound Logistics: Smart Distribution and Fulfilment. AI is reshaping how finished products reach customers across every link of the outbound chain. In warehouse automation, e-commerce operators use AI to generate optimal layouts and picking routes, producing visual heat maps of high-traffic areas and suggesting quarterly reorganisations. Shipment documentation, labels, customs declarations, certificates of origin, and compliance paperwork for international shipments, is generated automatically, cutting processing time from hours to minutes. Last-mile optimisation considers package size, delivery windows, traffic patterns, and driver capabilities to create efficient routes, complete with turn-by-turn instructions and customer-communication templates. When customers initiate returns, AI generates personalised return labels, analyses return reasons to flag emerging quality trends, and creates restocking or disposal recommendations. And across distribution networks, AI maintains real-time inventory visibility, producing status reports and triggering alerts and recommended actions whenever stock levels cross replenishment thresholds.

4. Marketing and Sales: Personalisation at Scale. The most visible transformation happens in customer-facing activities. Content generation reaches a level of variety that was previously infeasible, a consumer-goods company can produce 10,000+ product-description variations optimised for different demographics, channels (web, mobile, print), and cultural contexts while keeping a single brand voice intact. Campaign creative follows the same pattern: marketing teams generate hundreds of ad variations (headlines, body copy, visuals, CTAs) for A/B testing across platforms, and the winners are chosen by measured response rather than by the loudest voice in the creative review. In sales enablement, AI generates personalised presentations, proposals, and ROI calculators based on prospect industry, company size, and pain points surfaced during discovery calls. Market research shifts from periodic studies to continuous synthesis, with AI analysing thousands of customer reviews, social-media conversations, and competitor websites to produce intelligence reports, trend analyses, and strategic recommendations. Dynamic pricing strategy models combine competitor prices, inventory levels, seasonal demand, and customer purchase history into recommendations and accompanying rationale documents for sales teams. And lead nurturing is increasingly choreographed by AI, which creates personalised email sequences, follow-up messages, and content recommendations and adjusts tone and timing based on each prospect's engagement.

5. Service: AI-Augmented Customer Support. Post-sale service becomes both more efficient and more personalised when AI is layered into every step. Customer-support automation handles the routine majority of inquiries (the share depends entirely on how well the knowledge base behind it is maintained), and when something more complex surfaces, the bot generates a comprehensive case summary for human agents that includes the customer's history, previous interactions, and suggested resolution paths. Troubleshooting guides for new product issues, step-by-step instructions with screenshots and videos in multiple languages, can be generated within hours rather than days. In warranty processing, AI reviews claims, drafts approval or denial letters with clear explanations, and produces fraud-pattern reports that flag suspicious clusters. Knowledge-base maintenance becomes continuous: as tickets are resolved, AI automatically updates articles, creates FAQ entries, and generates training materials for support staff. And in customer-feedback analysis, AI synthesises support tickets, surveys, and social-media mentions into monthly satisfaction reports with prioritised, actionable improvement recommendations.

Support Activities: AI Enabling Organizational Excellence

Firm infrastructure (accounting, finance, legal, planning). AI is reshaping the back office in much the same way. Financial reporting can be generated end-to-end from raw transaction data, comprehensive reports, variance analyses, and executive summaries paired with visualisations and narrative explanations of key metrics. Compliance documentation becomes self-maintaining as regulatory AI generates required filings, audit trails, and policies, refreshing materials whenever rules change. In strategic planning, AI assists scenario work by producing financial projections under different market conditions and assembling detailed planning documents with built-in risk assessments. And in contract management, legal AI scans contracts for standard clauses, generates redlines, summarises key terms, and flags non-standard provisions for attorney review.

Human resource management. Recruitment teams use AI to draft job descriptions optimised for different platforms, write personalised candidate outreach, and prepare interview question sets tailored to role requirements and company culture. Onboarding becomes more personal at scale: AI builds bespoke schedules, role-specific training materials, and welcome packets customised to department and location. In performance management, AI analyses performance data to generate review summaries, development-plan recommendations, and career-path suggestions, helping managers provide more structured feedback. And in learning and development, training AI assembles customised learning paths, course materials, and knowledge assessments based on role requirements and identified skill gaps.

Research and development. Literature review shifts from a multi-week chore to a continuous capability as research AI analyses thousands of scientific papers to identify knowledge gaps and research opportunities. Experimental design AI produces protocols, statistical-analysis plans, and safety documentation for new initiatives. In patent research, IP AI conducts landscape analyses, generates prior-art reports, and even drafts patent applications from invention disclosures. Technical documentation closes the loop, turning research findings into specifications, white papers, and regulatory submission documents.

Procurement. AI gives buyers a clearer view across the supply base. Vendor analysis produces comprehensive comparison reports covering pricing, quality metrics, delivery performance, and risk factors. RFP creation generates detailed proposal documents, evaluation criteria, and vendor scorecards in a fraction of the usual time. Contract negotiation is supported by AI-generated strategies based on market analysis, counteroffer templates, and clause-level comparison documents. And spend analysis turns procurement data into insights on spending patterns, cost-saving opportunities, and strategic sourcing recommendations.

The Integrated Advantage: The real power of the GenAI value chain emerges when these capabilities work in concert. Demand forecasts from inbound logistics feed production planning in operations, which guides marketing campaign timing, which shapes how service teams staff up. Optimizing individual functions is only the start; the larger prize is a business where insights flow across the old functional boundaries instead of stopping at them.

8.2 Marketing: Transforming Content and Customer Engagement with AI

Marketing adopted generative AI earlier, and more aggressively, than almost any other function. The reason is simple: marketing runs on content, and generative AI produces content. It drafts copy, generates images and video, and tailors messaging to different audiences while keeping the brand voice intact. Picture a team producing fifty product descriptions before lunch instead of five. That is the value proposition: faster production, lower creative costs, and campaigns personalized in ways that used to be unaffordable.

The Shift to Vibe Marketing

Brands now face a mandate to rethink marketing, moving along a spectrum that runs from vibe marketing to fully agentic marketing. What changes along that spectrum is how a campaign travels from initial insight to final execution.

But what exactly is vibe marketing? The term spread through practitioner circles in 2025 (an extension of Andrej Karpathy's "vibe coding" coinage), and as used in this book it means: the practice of using GenAI access, assistants, applications, automation, and agents to execute large-scale campaigns so that human marketers can remain entirely focused on strategy, taste, and "the vibe".

By offloading the heavy lifting of execution to AI, marketers can reclaim their creative bandwidth. This process is generally structured across three distinct phases.

Phase 1: Preparation and Insights

To capture the right vibe, marketers must first understand their audience at scale. AI enables the rapid and cost-effective scaling of qualitative customer research.

• Marketers can utilize AI-moderated interviews that feature multi-modality and multi-language support.
• Teams can uncover insights by analyzing "dark data," such as call recordings that reveal customer anxiety, confusion, and unmet needs.
• Chat logs can be processed to highlight customer friction in real time.
• Customer complaints can be analyzed to reveal emerging risk and trust signals.

Phase 2: Augmented Creation

With insights secured, AI acts as a creative multiplier, allowing marketers to focus on curating the best outputs.

• This phase allows for massive scaling and variation of visual assets specifically for ideation.
• It enables content remixing, intelligent drafting, and rapid prototyping.
• Brands can employ "synthetic consumers" or digital twins to test 1,000 message variants before a campaign launches.
• Marketers can simulate reactions based on specific personas, life stages, and need states.
• Synthetic focus groups help teams identify likely customer objections before a formal compliance review.

Phase 3: Execution and Distribution

The final phase uses AI to push the curated "vibe" into the market efficiently through Generative Engine Optimization (GEO) and Agentic Marketing.

• GEO requires a shift away from traditional search engines, forcing marketers to approach optimization more like PR and earned media.
• This approach prioritizes brand impressions over traditional clicks and adapts to the rise of chatbot advertising.
• Agentic marketing allows brands to orchestrate a team of specialized AI agents.
• These agents can automatically monitor the market for competitor product launches.
• Agents can also handle predictive media buying, allocation, and automated performance tracking.

The Authenticity Imperative

While AI provides the scale and speed necessary for this framework, there is a distinct danger to the "vibe" if human oversight is removed. Fully AI-generated marketing content is often compared to "plastic flowers". To ensure the final campaign resonates with real human emotion, marketers must adhere to a few core principles:

• Treat AI outputs as inputs: AI-generated content should serve as the raw material for human-centric content creation, not the finished product.
• Be authentic: Brands must maintain their unique voice.
• Tell stories: The focus should remain on genuine narrative and human connection.

Generative AI for Content Creation and Campaigns

Brands are using AI to generate marketing content that used to require weeks of human creative effort. Microsoft, for example, used generative AI tools to script, storyboard, and produce a video ad for its Surface devices, cutting production time and cost by 90%. Viewers did not even notice the ad's AI-generated elements. Quality survived the speed.

Coca-Cola's holiday campaigns are the most public test of the same idea, and the most instructive. In November 2024 the company released fully AI-generated remakes of its 1995 "Holidays Are Coming" commercial, produced by three AI studios in a fraction of the usual time, and drew loud criticism for uncanny visuals and for using AI at all on a beloved piece of brand heritage. Rather than retreat, Coca-Cola returned in November 2025 with a new AI-generated version made by a small team with far better tools, generating tens of thousands of clips to select a few dozen, and the reception was markedly warmer. The two-year arc is the lesson: the technology improved faster than the critics expected, the brand risk was real but survivable, and the human choices, which story to tell and which clips to keep, were where the quality came from. These projects show that AI can shoulder much of the heavy lifting in content generation, provided human marketers stay in charge of taste and strategy.

Personalization at Scale

A major promise of AI in marketing is the ability to deliver personalized content and customer experiences at scale. Generative AI can dynamically create product descriptions, emails, or ads tailored to individual consumer segments, something that would be infeasible to do manually for millions of customers. The marketplaces show what this looks like at scale. Amazon's generative listing tools, rolled out from 2023 and expanded through 2025, let a seller produce a complete product listing from a few words, a product URL, or an image, and hundreds of thousands of sellers had adopted them within a year; the same models write the review summaries shoppers see on every product page and power Rufus, the shopping assistant that answers questions in the seller's stead. Alibaba, which pioneered machine-written product copy for its merchants years earlier, now offers the same through its Qwen models to sellers across its platforms. The pattern in both cases is a marketplace using generative AI to raise the floor of content quality for millions of small businesses at once.

This is what mass personalization looks like in practice: each customer sees content crafted for them, whether a localized campaign variant or a dynamically written product summary, and each seller gets copy that once required an agency.

Enhancing Creativity and Branding

Beyond efficiency, generative AI can also expand creative horizons for marketing teams, and by 2025 it was changing the economics of the most expensive creative format of all, the television commercial. In June 2025 the prediction-market company Kalshi aired a thirty-second spot during the NBA Finals that was generated entirely with Google's Veo 3 video model. One filmmaker made it alone in two days for about two thousand dollars, generating several hundred clips and keeping fifteen; a comparable conventional production would have cost hundreds of thousands of dollars and taken months. The ad was deliberately absurd, which suited the tool's limitations, and it worked because the creative judgment, the joke and the edit, was human. It is the clearest single demonstration that video production cost is no longer the constraint on brand creativity; taste is.

At the other end of the scale, Unilever rebuilt product photography itself. In 2025 the company began creating digital twins of its products, single files holding every variant, label, package, and language, and generating imagery and video from them with real-time 3D rendering and generative AI in place of physical shoots. Piloted in its Beauty and Wellbeing group for TRESemmé, Dove, Vaseline, and Clear, the approach reported up to 55 percent lower content costs and 65 percent faster turnaround, with higher click-through and attention in market, and it cut a five-step content workflow to one. Cost is only half the story. In both cases the AI worked as a creative collaborator, and in both the brand kept a human deciding what was on-brand and what was not.

Real-World Case Snapshots

Table 8.1 highlights several real-world applications of generative AI in marketing across different industries and regions, along with their outcomes. Dates are given because this market moves fast: the 2024 entries are already history, and the 2025 ones will be by the time you read this.

As the table shows, brands across the US, Europe, and Asia are putting generative AI to work in marketing, and they report faster content cycles, higher campaign ROI, and new ways to engage customers. A 2024 marketing industry survey found that 51% of marketers have used or plan to use generative AI, with image generation (69% of marketers) and text generation (58%) the most common applications. The platform ecosystem is broad: Jasper and Copy.ai for copywriting, Midjourney and Adobe Firefly for images, Synthesia and HeyGen for synthetic video ads, and the leading large language models for everything from social media posts to strategy briefs. The practical advice is to pick tools that fit your brand and keep a human reviewing the output for voice, accuracy, and bias. Used that way, generative AI is a genuine co-creator rather than a liability.

8.3 R&D: Accelerating Innovation and Design

Beyond marketing, generative AI is reshaping Research & Development (R&D), the engine of innovation for products and services. R&D spans everything from designing physical products and developing new formulas to writing software and creating entertainment content. Where generative AI earns its keep is speed of exploration: it generates prototypes, simulates ideas, and optimizes designs far faster than traditional methods. An engineer can have it explore thousands of possibilities, whether design permutations, molecular structures, or code solutions, in the time it once took to test a handful. Development cycles shrink accordingly.

Product Design and Engineering

One of the most tangible impacts of generative AI is in product design, often via generative design software that creates optimized geometries under given constraints. Two cases from the late 2010s remain the clearest physical illustrations, and they predate the language-model era: generative design is an older cousin of today's generative AI, and it is included here because the logic, explore a vast design space and let the human choose, is identical. A classic example is General Motors (GM), which partnered with Autodesk to redesign a simple but critical part: the seat-belt bracket that anchors seat belts in cars. Using generative design algorithms, GM's engineers input the requirements (attachment points, load forces, allowable materials, etc.), and the software produced over 150 alternative designs, many with organic, complex shapes a human alone might never conceive. The chosen AI-generated design combined what used to be eight separate components into a single 3D-printed part. Remarkably, the new bracket is 40% lighter and 20% stronger than the original design, contributing to vehicle weight reduction (which matters greatly for electric cars' range and efficiency) without sacrificing strength.

In Europe, Airbus applied a similar approach to an interior aircraft component, a partition wall. Using Autodesk's generative design, Airbus created a "bionic partition" with a nature-inspired lattice structure that cut weight by 45% while maintaining full strength. Lighter airplane parts mean real fuel savings and lower emissions. In automotive and aerospace R&D alike, generative AI is producing lighter and stronger designs while compressing months of design iterations into weeks.

Faster Product Development Cycles

Generative AI is also helping companies dramatically speed up the R&D process for new products. In the fast-moving consumer goods (FMCG) sector, PepsiCo provides a striking example. Traditionally, developing a new snack flavor or product variation might take 6 to 12 months of brainstorming, formulation, and testing. PepsiCo's R&D teams turned to generative AI to accelerate this. By analyzing vast combinations of ingredients, flavors, and consumer taste data, AI can suggest optimal recipes and even novel snack shapes. Using these tools, PepsiCo managed to shrink the development cycle for a new Cheetos snack to just six weeks, a fraction of the usual time.

"Generative AI allows us to optimize product attributes in a way that was previously unimaginable," says Athina Kanioura, PepsiCo's Chief Strategy and Transformation Officer, noting that AI helped reduce the number of test cycles needed for the new Cheetos product. The generative models proposed flavor variations and ingredient mixes that met consumer preferences for taste and texture, which R&D could then quickly prototype. Beyond new products, PepsiCo also uses AI to reformulate existing snacks (like lowering sodium or fat) by simulating ingredient interactions, delivering healthier options without extensive trial-and-error in labs. This data-driven innovation ensures that product launches are both faster and more aligned with consumer trends, giving companies a competitive edge in crowded markets.

Scientific Discovery and Pharma R&D

In more research-intensive fields such as pharmaceuticals and materials science, generative AI is breaking new ground by designing entirely new molecular structures and compounds, and 2025 brought the first clinical proof. Insilico Medicine first made news in 2019 when its generative model designed candidate molecules against a fibrosis target in 21 days, a process that typically takes months. The company then used its generative platform to discover a new target for idiopathic pulmonary fibrosis and design a drug against it, rentosertib, and in June 2025 Nature Medicine published the results of a randomized Phase IIa trial: the first time a drug with both an AI-discovered target and an AI-designed molecule had shown a positive efficacy signal in patients. At the highest dose, lung function improved on average where the placebo group declined, and the company moved the drug into Phase III. One trial is not a cure, and the hard, expensive work of late-stage development remains. But the question of whether generative AI can produce a real drug candidate that behaves in humans has now been answered in the affirmative.

Traditionally, drug discovery is like searching for a needle in a haystack, screening countless molecules to find a few hits. Generative AI flips this paradigm by creatively suggesting molecular structures that fit the desired criteria (shape, chemical properties, target binding) without brute-force testing of each option. Pharmaceutical giants and startups alike are now using such AI platforms to generate ideas for new drugs (for cancer, antibiotics, etc.), cutting the search time from years to days in some cases. Human chemists still must synthesize and validate the AI's suggestions, but the efficiency gains upstream are enormous.

Software and Content Development

Generative AI also boosts R&D productivity in software engineering and digital product development. AI coding assistants, the best-known being GitHub Copilot (now powered by the latest GPT and Claude models, and joined by fully agentic tools like Claude Code and Codex; see Chapter 7), help developers generate code from natural-language prompts. This has sped up programming tasks significantly: GitHub's controlled study found developers completed a benchmark task 55.8% faster with Copilot. In practice, this means feature prototypes can be built in days instead of weeks. The largest measured deployment is in maintenance rather than invention: in 2024 Amazon used its Q Developer agent to upgrade more than 30,000 internal Java applications to a modern version, cutting the average upgrade from about fifty developer-days to a few hours, which the company estimated saved 4,500 developer-years of work and about $260 million a year in infrastructure cost. That is the pattern Chapter 15 describes: agents excel where the goal is precise and verification is mechanical, and migration backlogs are exactly that.

Beyond coding, generative AI can create synthetic data (for testing or training models), design user interface layouts, or even generate game assets and levels in the entertainment industry. Game developers, for instance, are exploring AI tools that generate virtual environments and characters, speeding up creative R&D for new titles. The common theme across all of these applications: generative AI explores a solution space at high speed, whether that space holds physical designs, chemical compounds, or lines of code, and presents the most viable options to human experts.

Real-world Impact

Companies adopting generative AI in R&D report tangible benefits: shorter time-to-market, improved product performance, and greater innovation output. A McKinsey analysis notes that leading firms are using AI to "size potential markets, analyze competitor moves, and estimate the value of different strategic initiatives across multiple scenarios" during product strategy and design. The R&D function, once reliant purely on human trial-and-error and experience, is becoming more data-driven and simulation-driven. Human expertise has not gone anywhere: engineers and scientists still set the goals, validate the results, and supply the contextual judgment that AI lacks. What changes is the risk profile of innovation itself. Teams can test far more ideas virtually, catch failures early, and double down on winners, whether they are designing the next electric vehicle or formulating a new cosmetic.

8.4 HRM: AI for Talent, Hiring, and Employee Support

Human Resource Management might seem like a domain focused on people, recruiting them, developing them, engaging them, and it is. It is also a domain full of repetitive, communication-heavy work, which makes it ripe for intelligent automation. Generative AI is helping organizations attract talent, streamline hiring, personalize training, and improve employee self-service. It handles the grind (writing job descriptions, answering the same benefits question for the hundredth time) and surfaces insights buried in HR data, from resume screening to career development planning. For a time-pressed HR team, that means more hours for the interpersonal work that actually requires a human.

Talent Sourcing and Recruitment

Finding the right candidates is a perennial challenge. Generative AI can improve this through intelligent resume screening, automated outreach, and even drafting job postings that attract a broader talent pool. For example, RingCentral, a cloud communications company, faced slow, manual processes in sourcing specialized talent. By working with an AI talent platform (Findem), RingCentral was able to automatically comb through 1.6 trillion data points from internal and external sources to identify candidates that matched very specific criteria. The AI not only matched resumes to job requirements but also drafted personalized outreach messages. The result was a 40% increase in the candidate pipeline and a 22% improvement in candidate quality, including a 40% boost in applicants from under-represented groups.

Platforms like LinkedIn have built on generative AI's knack for understanding job requirements and describing them well. In 2023, LinkedIn introduced AI-powered tools that let recruiters enter a few key details and have GPT-3.5 generate a polished job description draft. Recruiters save time (LinkedIn noted the tool "does the heavy lifting" in composing the text), and the language tends to come out more inclusive and appealing, which widens the candidate pool.

In fact, T-Mobile used an AI writing assistant (Textio) to review and enhance the wording of its job postings and recruiting emails, to mitigate unconscious bias and improve diversity in hiring. By integrating the tool into their workflow and even into their Workday applicant tracking system, T-Mobile's recruiters got real-time suggestions for more inclusive language, helping the company make faster progress on its diversity goals. These examples show AI's dual benefit in recruitment: efficiency gains (faster sourcing and communication) and higher quality outcomes (more diverse, well-matched candidates).

Streamlining Hiring Processes

Once candidates are in the pipeline, generative AI can continue to optimize the hiring funnel. Mastercard, for instance, dealt with an enormous volume of applications as the company expanded. They implemented an AI-driven talent platform (Phenom) that automates parts of the hiring journey, from scheduling interviews (using AI to match calendars) to answering candidate FAQs via chatbots. The impact was dramatic: interview scheduling was 85% faster, with 88% of interviews booked within 24 hours of the request. And by enhancing their career site with AI (including features for candidates to join talent communities and receive tailored job recommendations), Mastercard saw a 900% increase in its candidate database and ultimately added 2,000+ new hires sourced through these AI-powered channels. What would have required a much larger recruiting team (to manually chase schedules and follow up with hundreds of thousands of applicants) was accomplished with intelligent automation.

Another growing practice is using AI to conduct preliminary candidate assessments: for example, AI video interview platforms that pose questions and evaluate responses. A cautionary tale is instructive here: several large employers experimented in the late 2010s with AI video interviews that analyzed candidates' word choices and even facial expressions. The facial-analysis component was abandoned by its leading vendor in 2021 under bias criticism, and emotion inference in hiring now sits squarely in the EU AI Act's restricted zone (Chapter 13). The episode is the pattern to remember: an HR use case can be technically impressive, efficient, and legally untenable at the same time, so screening tools must clear the governance bar before the efficiency bar. Generative AI takes this further by potentially simulating "role-play" interviews or crafting custom interview questions on the fly based on a candidate's resume. While companies must be cautious and ensure fairness (to avoid bias in AI decisions), these tools can make hiring faster, fairer, and more engaging for candidates.

Employee Self-Service and HR Support

HR departments field endless employee questions, about benefits, payroll, policies, or IT issues, which can overwhelm HR staff. Generative AI-powered HR assistants can handle a large portion of these routine queries instantly and accurately, and the most consequential example is IBM's. Its AskHR agent, built up over several years and rebuilt on generative models, was by 2025 automating 94 percent of routine HR tasks, from pay statements to leave questions, and in May 2025 IBM's chief executive, Arvind Krishna, said publicly that AI had taken over the work of several hundred HR employees. The rest of the story matters as much as the headline: IBM's total headcount grew, because the company redirected the savings into hiring programmers and salespeople, roles it described as facing other humans rather than doing rote process, and its parallel AskIT agent cut calls and chats to the IT help desk by 70 percent. The case is the cleanest illustration in this book of what automating a support function actually does: it does not remove the function, it moves the people to work that needs judgment, and it makes the accountability for that move a leadership decision rather than a technology outcome.

Smaller organizations follow the same pattern with narrower assistants: Straits Interactive, a data-governance firm in Singapore, built a generative AI assistant that interprets privacy regulation in plain language so employees get compliance guidance without waiting for an expert. In large organizations, each employee now plausibly has a personal HR advisor available around the clock, for policy questions and, increasingly, for career advice ("What training should I take to be eligible for promotion?"). This level of personalization at scale was never feasible before, and it is high-touch support delivered through high-tech means.

Learning and Development

Another HR area seeing generative AI impact is employee training and development. AI can generate customized learning content, like training manuals, quiz questions, or even simulated role-play scenarios for practice. For example, a sales team could use a generative AI tool to simulate customer interactions (the AI plays the role of a difficult customer, allowing the employee to practice responses). There are already platforms where AI generates coaching tips or performance review drafts based on an employee's achievements. LinkedIn's generative AI features extend to LinkedIn Learning, where new courses on AI are being added and AI might eventually tailor learning paths for users. In performance management, managers are using AI to help draft more effective and unbiased performance evaluations by analyzing an employee's contributions and writing a first pass of feedback for the manager to refine. These uses are still emerging, but they hint at a future where AI plays a supportive role throughout the employee lifecycle, from hire to retire.

HRM Use Cases and Outcomes

Table 8.2 summarizes a few real-world generative AI applications in HR and their results, with dates, since the older recruiting cases predate generative models and the IBM case shows where the function is heading:

These examples underline that HR is becoming more data- and AI-driven. A caution is in order, though. Fairness, transparency, and ethics carry particular weight in HR: an AI model trained on biased historical data can quietly favor or reject candidates for the wrong reasons. The HR leaders I speak with are careful to keep humans in the loop, using AI to assist rather than to decide hires or promotions outright. Done right, generative AI can actually make HR processes more human, freeing time for relationship-building and coaching while giving each employee or candidate personalized support. As one HR executive put it, AI can handle the grunt work, letting HR professionals "be more strategic and enhance their impact". From global corporations to startups, HR teams across the US, Europe, and Asia are piloting these tools to compete for talent and improve the employee experience.

8.5 General Management and Strategy: Decision Support, Productivity, and Innovation

Generative AI has also found its way into the executive suite. Business leaders and strategists are using it to analyze complex data, generate insights, support decisions, and even sketch strategy options. In day-to-day management, AI tools draft reports, summarize market research, prepare presentations, and speed up communication. For strategic planning, AI can produce scenario analyses in minutes or comb through competitive intelligence to suggest moves. The net effect: executives and managers make better-informed decisions faster, with AI-generated knowledge and recommendations at hand.

Enhanced Decision Support and Analysis

One of the challenges in management is digesting overwhelming amounts of information, financial reports, market research, news, internal data, to make timely decisions. Generative AI (particularly the large language models) excels at summarizing and synthesizing information. Managers can ask an AI assistant to "Summarize the key trends in our quarterly sales report" or "Give me a SWOT analysis based on these 50 pages of market research". The AI will generate a concise briefing, saving hours of reading.

For example, wealth management firm Morgan Stanley built an OpenAI-powered assistant that allows its financial advisors to query a vast library of research reports and get instant, distilled answers. Advisors can ask, say, "What's our latest outlook on European tech stocks?" and the AI will pull from internal research and generate a helpful summary or even draft an email to clients with the key points. This tool, called AI @ Morgan Stanley Assistant (with a feature named "Debrief"), automatically creates meeting summaries and follow-up notes, potentially saving thousands of hours for their 15,000 financial advisors.

The principle extends to general management: AI can act as an omnipresent analyst, crunching numbers and reading documents to answer managers' ad-hoc questions. Microsoft's Power BI, a business intelligence tool, has integrated generative AI (the new Copilot feature) that lets managers use natural language to probe data and generate visuals or insights. This means a sales manager could simply type, "Show me a chart of Q3 revenue by product line and identify the top 3 growth drivers," and the AI will produce the chart and a brief analysis, tasks that previously required a dedicated analyst. By doing in minutes what used to take days, generative AI is helping managers be more agile and data-driven.

Strategic Planning and Scenario Generation

Crafting business strategy often involves asking "what if" and considering multiple future scenarios. Traditionally, scenario planning is a time-consuming exercise done by strategy teams. Generative AI can radically speed this up by generating detailed scenario narratives and even quantitatively modeling different assumptions. According to Harvard Business Review, "GenAI can help organizations overcome inherent shortcomings in conventional processes for performing contingency scenario planning".

For instance, a strategy team can prompt an AI with something like: "Imagine three scenarios for our industry in five years: one optimistic, one moderate, one pessimistic. Describe each and the potential strategic moves we should make." The AI can produce rich narratives and even suggest strategic initiatives for each scenario. While these AI-generated scenarios are starting points (human strategists will refine and stress-test them), they significantly cut down the time needed to explore strategic options.

Consulting firms are already using such approaches with clients. Bain & Company, a leading global consultancy, formed a strategic alliance with OpenAI to embed GPT-4 into its consulting services. In one high-profile project with Coca-Cola, Bain used generative AI to help brainstorm and develop new marketing strategies and content (the Coca-Cola "Create Real Magic" campaign was an outcome of this collaboration). Beyond marketing, Bain reported that generative AI is being used to simulate market entry strategies and to generate draft strategy documents that consultants and clients can then iterate on.

Similarly, McKinsey & Company has observed that AI can "enhance every phase of strategy development, from design through execution", allowing organizations to analyze competitors, size markets, and estimate the value of strategic moves far more quickly than before. For example, if a company is considering a new product launch, AI can swiftly summarize all competitor products, patent filings, and consumer reviews in that space, giving strategists a high-level view to base their plans on. Leaders can explore more ideas, with more evidence behind each one, before committing.

Productivity and Collaboration Tools for Management

On a day-to-day basis, generative AI is becoming like an executive assistant that never sleeps. Consider the flood of emails, memos, and documents a typical manager deals with. AI-powered tools (like Microsoft 365 Copilot and Gemini for Google Workspace) can draft responses to emails, create first drafts of documents, and even build slide presentations from a simple outline. For instance, a manager could ask, "Draft a project update email highlighting our progress and next steps," and the AI will generate a polished email which the manager can then tweak and send. Equally, if a regional manager wants a PowerPoint on last quarter's performance, an AI can generate slides with charts and bullet points drawn from the raw data.

Global professional services firms are rapidly embracing these capabilities internally. PricewaterhouseCoopers (PwC), for example, made a $1 billion investment in AI and has become one of the largest enterprise users of ChatGPT. In 2024 PwC announced it would roll out ChatGPT Enterprise to its 75,000 U.S. and 26,000 UK employees as a productivity tool. PwC is even developing custom AI models ("private GPTs") for tasks like reviewing tax documents or generating financial reports. The goal is for consultants and auditors to complete analysis and documentation in a fraction of the time. "We are actively engaged in genAI with over 95% of UK and US consulting client accounts," PwC noted, highlighting how ubiquitous they expect AI usage to be across their business lines.

Banking offers the largest measured rollout to date. JPMorgan Chase launched LLM Suite, a proprietary generative AI platform, in the summer of 2024 and had more than 200,000 employees on it by mid-2025, from legal and sales to client services; the bank reports one to two hours saved per user per week, and, more candidly than most, that only about half of the employees with access use it on a given day, which is the adoption paradox of Chapter 2 in a single number. American Banker named it the 2025 Innovation of the Year. The lesson executives take from it is not the tool but the operating model: a central platform, hundreds of use cases prioritized against measured value, and honest reporting of who actually uses what.

Another Big Four firm, KPMG, is integrating generative AI to assist its legal and advisory teams in drafting documents (e.g. contract analysis) and in internal knowledge management. These large-scale adoptions underscore that AI is becoming a standard part of the managerial toolkit, as common as spreadsheets or email. Managers in Asia are on the same trend; for instance, Japan's banking and telecom sectors are experimenting with bilingual AI chatbots to assist managers in summarizing English reports into Japanese and vice versa, bridging language gaps in global operations.

Corporate Strategy and Innovation

Many companies view generative AI not only as a tool for current managers but as a strategic asset for the business model itself. For example, consulting firms are productizing generative AI solutions for their clients (as seen with Bain & OpenAI, or Deloitte's AI practice building domain-specific GPT models). In the tech industry, Salesforce rebuilt its product line around Agentforce, agents that handle service, sales, and marketing tasks inside the CRM, and used its own deployment as the reference case: by 2025 the company said agents were resolving a large share of its customer-support conversations and it had slowed hiring in the affected teams while adding sales roles to sell the agents. These strategic AI implementations often come from the top: executive leadership needs to champion and invest in AI capabilities across the organization.

The adoption curve has been steep: 72% of companies worldwide were using AI in at least one business function by early 2024, and by late 2025 the figure had reached 88% (McKinsey State of AI waves; see Chapter 2 for the full series and its paradoxes). Companies are forming AI centers of excellence and training programs to raise AI fluency among managers. In Europe, for example, Siemens and Bosch have launched internal initiatives to have their managers pilot generative AI in operations and strategy, sharing best practices and success stories. The competitive advantage of using AI at the management level can be substantial: decisions get made faster and with more evidence; strategies are tested virtually before committing real resources; and the organization becomes more adaptive through continuous learning from AI-generated feedback.

In embracing generative AI for management and strategy, companies should also develop proper governance. AI-generated content can sometimes be inaccurate or might reflect biases in its training data. Therefore, leading organizations pair AI outputs with human review, a manager might use the AI's analysis but will cross-verify key facts, or use an AI's suggested strategy as one input among others (including human intuition and stakeholder consultation). When used wisely, generative AI can augment managerial judgment, not replace it. It serves as a kind of "co-pilot" for executives: crunching data, offering second opinions, proposing creative solutions, and taking over routine chores. This frees leaders to do what they do best, provide vision, empathize with employees and customers, and make the tough calls that machines alone cannot.

8.6 Conclusion: From Hype to Real Value Across the Business

Generative AI is already delivering real value across the business functions we have covered. In Marketing it powers personalized campaigns and prolific content creation, and has collapsed the cost of video. In R&D it shortens innovation cycles and cracks design problems, from a lighter airplane part to a drug candidate that has now shown efficacy in patients. In HRM it streamlines hiring and supports employees by taking over repetitive tasks. In General Management and Strategy it helps leaders analyze information and formulate plans with more speed and insight. None of this is theoretical. The case studies cited in this chapter come from working companies in the U.S., Europe, and Asia, supported by a growing ecosystem of AI tools and platforms.

The rewards come with a caveat: implementation needs thoughtful change management. The companies that have succeeded, from PepsiCo to JPMorgan, typically started with pilot projects, took data security and ethical use seriously, and trained their teams to work alongside AI. This chapter has stayed deliberately practical and non-technical because the aim is for managers in any function to see how generative AI can help them and what others have achieved with it. The same instrument that writes a marketing tagline one minute can help draft a strategic plan the next.

Generative AI is becoming deeply embedded in how companies operate and compete. Marketers find it a creative partner. R&D teams see an innovation catalyst. HR leaders view it as a co-worker that handles routine queries, and executives treat it as an ever-ready advisor. The technology is still evolving, and new applications will keep emerging beyond the functions covered here (AI in finance for fraud detection, say, or in supply chain for dynamic route optimization, which other chapters touch on). But generative AI has already moved from concept to pilot to broad adoption in many enterprises, reshaping job roles and required skills along the way. AI can generate content and ideas; human judgment, creativity, and empathy remain irreplaceable. The organizations that blend the two well are already collecting the returns.

As you look at your own business or functional area, the question is no longer "Should we use generative AI?" It is "Where, specifically, will it help us most?" Treat the case studies in this chapter as a starting menu of proven answers.