Appendix A
Generative AI by industries
This Infosys survey covers 12 sectors across Australia, China, India, Japan, New Zealand, and Singapore , allowing us to compare how extensively and effectively each industry is using generative AI.
Although most companies are investing in this technology, there is a wide range of spending among sectors.
While there are small differences between most industries in the frequency of implementation, in some industries this difference is significant. Automotive, energy, mining and utilities, and high-tech industries report significantly higher implementation of generative AI use cases. There are also statistically significant differences in industries that have implemented use cases that create business value.
We find that financial services, high-tech, insurance, and manufacturing report having significantly more initiatives that create value than almost all of the other industries in the survey.
Healthcare is the only industry that significantly lags other industries in both implementation and creating business value.
Appendix A: Generative AI by industries
Figure 14. Generative AI adoption and business value by industry
Source: Infosys Knowledge Institute
Generative AI Radar: APAC
Product design is a key focus for automotive companies, so it is not unexpected that nearly half of the APAC firms in our survey expect streamlined product development and design to be the most impactful use case for generative AI.
Automotive: Most positive impact expected from generative AI
Potential use cases
Quality control — Image processing to identify production defects in parts and alert staff so they can resolve the problems
Inventory management — Predict client demand for specific vehicles which can lead to optimal production and inventory levels
Contract management — Review and summarize supplier and dealer contracts to optimize relationship management
Autonomous vehicle training — Create virtual environments and synthetic data for realistic simulations
User personalization — Power in-car personal assistants, and predict preferred routes and dashboard settings
Road freight — Plan effective and efficient delivery routes
As might be expected for companies in the region that focus on winning consumers over with marketing content, nearly two thirds said they expect improved content generation and creativity to be the most impactful use case for generative AI.
CPG: Expectations of most positive impact from generative AI
Quality control — Identify defects in products and packaging, and issue automatic order refills
Product design and packaging — Analyze consumer preferences, trends, and social media to inform new product and packaging designs
eLearning content — Analyze consumer information such as customer service call and chats to develop learning materials for new products
QR codes as digital art — Transform QR codes into visually appealing brand identifiers
Consumer research — Generate synthetic customer data to test retail scenarios, and produce summaries and reports
Visual merchandising — Optimize planograms based on customer behavior and product data
Energy, mining and utilities companies in the Asia-Pacific region told us that they find operational efficiency, user experience, and product development to have the most impact when using generative AI technologies in their businesses.
Energy, mining, or utilities: Most positive impact expected from generative AI
Lease management — Use natural language processing to analyze leases for renewals, renegotiations, and cancellations
Site risk assessment — Analyze site safety measures and maintenance logs to predict equipment failures and downtime. This information can also be used to design response strategies
Refinery optimization — Process and analyze data to identify patterns and optimize processes in real time
Grid management — Analyze data on consumption patterns and load distribution to optimize grid performance and anticipate problems
Demand forecasting — Enhance accuracy of demand management, allowing for more efficient resource allocation and cost savings
Almost 40% of financial companies in APAC find the most positive impact of generative AI to be in creating new financial products, significantly ahead of the use case they consider the next most impactful, increased operational efficiency and automation.
Financial services: Most positive impact expected from generative AI
Claim processing — Analyze documentation and summarize relevant information such as coverage, incident reports and images, and supportingdocuments
Risk assessment — Continuously monitor market, news, and sentiment and generate reports to generate early warnings of events and mitigate risks
Streamline underwriting —Train models on document corpus to identify patterns to inform pricing and coverage recommendations
Fraud detection — Scan claims to identify anomalies and flag potential fraud
Customer service — Use chatbots built on LLMs trained with domain-specific data to manage customer interactions
Enhanced user exerience is the standout use case expected to have the most positive impact from generative AI among APAC's healthcare companies, with improved content generation and creativity seen as the least impactful potential use case.
Healthcare: Most positive impact expected from generative AI
Customer care optimization — Record, process, and analyze customer information, including medical records, calls, and emails, to save time in addressing customer problems and requests
Personalized care — Analyze human health data and speed up sequencing genomes, to identify health predispositions and develop personalized nutrition and care
Medical imaging — Improve the accuracy of medical imaging techniques, including CT and MRI, by automatically identifying anomalies
Patient triage — Use chatbots trained on specialized LLMs to provide first-line triage, assessing symptoms to send the patient to the appropriate human professional
As is the case with other industries, hightech companies in APAC see enhancing user experience as the most effective use case for generative AI, with 43% reporting this. The next most effective use case this industry sees is streamlined product development.
High tech: Most positive impact expected from generative AI
Quality assurance — Deploy image processing to quickly identify issues in website and application UX and UI so designers and developers can focus on solving the issue
Code management — Review code and create reports that can be used by engineers, business analysts, and product managers
Software development — Generate code, review existing code, and act as an assistant for developers
Business analysis — Create reports to provide insights to the entire business
Automation — Automate repetitive tasks, freeing up humans to do more creative or higher-level jobs
Product iteration — Speed up product development by generating and testing product ideas
In common with other companies and sectors in the region, insurance firms expect enhanced user experience to be the most effective use case for generative AI, although more firms here than in other sectors see content generation as an effective use case.
Insurance: Most positive impact expected from generative AI
Claim processing — Analyze documentation and summarize relevant information such as coverage, incident reports and images, and supporting documents
Portfolio management — Continuously monitor market, news, and sentiment to generate early indicators to better advise clients and manage banking assets
Data analytics — Generate reports to highlight anomalies
Document management — Use large language models trained on the firm’s data to summarize documents
Fraud detection — Identify anomalies and flag potential fraud
Customer service — Chatbots built on LLMs trained with domain-specific data can manage first-line interactions with customers, such as loan decisions
New product development is top of the list for life sciences firms in APAC when it comes to considering the most effective use case for generative AI, with the next most effective use case — increased operational efficiency — 25 points behind that top use case.
Life sciences: Most positive impact expected from generative AI
Synthetic data — Generate data to augment existing data to help improve the performance of Al models
Medical imaging — Improve the accuracy of medical imaging techniques by automatically identifying anomalies
Education — Use text-to-image Al to create images of body structures and processes to help students visualize anatomy and medical interventions
Drug discovery — Accelerate drug discovery by designing molecules and proteins with specific properties and optimize synthetic gene design forbiomanufacturing
Regulatory submissions — Summarize data needed for regulatory submissions and create drafts of required documents
Create novel molecules — Design drug-like molecules for specific use cases
Logistic and supply chain companies report that using generative AI will increase operational efficiency. They also expect to use generative AI for improved content generation, such as for drafting contracts and bidding documents etc.)
Logistics or supply chain: Most positive impact expected from generative AI
Maritime route optimization — Analyze vast amounts of data, such as currents, cargo weight, weather, and traffic, to plan the most optimal and fuel-efficient routes
Risk management — Analyze data on geopolitical concerns, weather, industrial unrest, and other factors to create dashboards that reveal insights and suggest mitigation
Supplier management — Chatbots can facilitate interactions with suppliers and create reports based on interactions and supplier data
Trends insight — Summarize information from external sources such as newspapers to understand trends that impact operations
Price intelligence — Gather information on competitor pricing and costs, and use chatbots to generate insights and reports
Streamlining product development and design is expected to be the use case with the most impact for this sector, with China leading in using generative AI for manufacturing. Enhanced user experience is second, 12 points behind product design.
Manufacturing: Most positive impact expected from generative AI
Quality control — Image processing to identify production defects in parts and alert staff to solve the problems
Continuous improvement — Integrating with IoT and edge computing to enhance data collection and analysis of manufacturing processes
Product design — Generate reports based on data from previous iterations of a product, including manufacturing issues, consumer preferences, and returns
Cost management — Reduce costs by streamlining maintenance and design processes
User feedback — Collate feedback reports from customers to generate recommendations for further refinement of products
Sentiments in retail and hospitality were statistically similar to the survey overall. About the same proportion expected generative AI to excel in each of the four areas of potential benefit.
Retail or hospitality: Most positive impact expected from generative AI
Personalized itinerary — Analyze customer search and purchase histories to generate tailored itineraries for customers before they book
Content generation — Create personalized content and campaigns for enhanced customer engagement
Data analysis — Speed up analysis of structured data, such as sales figures and emerging trends, to offer faster, deeper insights
Translation — Help visitors who don’t speak the local language to plan visits
Social listening — Monitor social media and generate reports based on user conversations and trends to fine-tune content and enhance personalization
Energy management — Analyze energy use patterns to save costs on hospitality and retail locations
In an industry that has a great deal of contact with customers and users, it is not surprising to see that firms here expect enhanced user experience and personalization to be the area where generative AI will have the most impact for them.
Telecommunications: Most positive impact expected from generative AI
Risk management — Automate responses to network irregularities through real-time data analysis
Customer support — Help users with billing queries and orders
Engineer support — Train generative AI on the network topology so that engineers can be guided through tasks with interactive assistance
Software development — Help developers build specialized code and applications more quickly
Network optimization — Analyze network data and conditions and generate insights to streamline resource development
Network security — Track threats and assess vulnerabilities by analyzing network traffic and generating reports to spot malicious activity
