rssitbuyer

AI Is Driving Infrastructure Spending as a Priority in Financial Services in 2026

Jerry Silva — Thu, 19 Mar 2026 04:00:00 GMT

This IDC Perspective describes the thinking of financial institutions that have made their intentions clear about the need for infrastructure improvements, and the inference IDC Financial Insights is making about AI as the main driver behind these decisions. Over three years have passed since GenAI made the news, and while the financial services industry struggled to create real benefits from generative AI, agentic AI and AI agents have further complicated the picture, and the industry, generally, seems to be unprepared to leverage these new technologies fully. But recent surveys of the industry by IDC point to a willingness (even a need) to invest in overcoming the challenges of inadequate infrastructure preparedness in 2026 and beyond.

"The financial services industry is in 'rebuild' mode in preparation for the AI-fueled business," said Jerry Silva, vice president, IDC Financial Insights. "IT investments in infrastructure modernization are outpacing investments in areas like customer experience, even though there isn't always a short-term lift in revenue."

AI Isn’t Going to “Eat” Software: Agentic AI Needs the Authoritative Data and Rules Inside Enterprise Apps

Heather Hershey, Mickey North Rizza — Thu, 19 Mar 2026 04:00:00 GMT

This IDC Perspective discusses enterprises’ adoption of AI agents within enterprise applications. Despite headlines suggesting AI will consume enterprise software, the evidence points to a more nuanced shift. As Jensen Huang noted at the Cisco AI Summit, the idea that AI replaces software misconstrues how computing works. IDC research shows enterprises expect agents to become a new intelligence layer across applications, yet alignment and security concerns reinforce continued reliance on established systems of record. The result is a transitional equilibrium in which agents orchestrate and interact, while core applications retain enforcement and accountability authority. IDC forecasts broad enhancement of enterprise applications by 2027 rather than their extinction.

“Stripped of hype, agentic AI is still software,” said Heather Hershey, research director, AI-Enable Digital Commerce at IDC. “Agentic AI is a probabilistic reasoning system that can pursue goals with limited supervision, yet it remains dependent on existing enterprise data foundations and governance frameworks. For executives, the strategic imperative is clear: capture value by embedding agentic AI within trusted application environments that preserve accountability and control.”

“‘Agents as apps’ is the new software model that will quicky enhance the software journey for organizations in the AI digital world,” said Mickey North Rizza, group vice president, Enterprise Software. “Invest in the future and do it now, but don’t throw the baby out with the bathwater. Your enterprise application software is a major foundation into the AI world — use it, mold it, and grow it.”

AWS and OpenAI To Expand Partnership With a Co-Developed Stateful Agentic Runtime

Matthew Flug — Thu, 19 Mar 2026 04:00:00 GMT

AWS and OpenAI announced a jointly developed stateful runtime environment for AI agents, delivered through Amazon Bedrock and designed to run natively within customers’ AWS environments. The runtime will enable AI agents to maintain persistent context, memory, workflow state, and tool connections across multi-step tasks. The goal is to simplify the development and management of AI agents by reducing the need for developers to manually write “glue code” or manage session history and orchestration logic. The new runtime integrates with AWS identity, security, and governance services, allowing organizations to implement existing enterprise controls and compliance boundaries into the runtime’s agentic workflows. The announcement underscores a shift in the application platform market from model-centric differentiation toward runtime architecture, governance integration, and vertically integrated AI infrastructure.

Data Sourcing and Pricing in the Age of AI: How Pricing, Rights, and Governance Are Shifting as AI Increases Data Demand and Scrutiny

Lynne Schneider — Thu, 19 Mar 2026 04:00:00 GMT

This IDC Perspective explores how AI is driving structural renegotiation of data pricing, governance, and monetization. Enterprises are distinguishing between training rights, retrieval access, and live connectivity models, each with distinct economic implications. Development-stage monetization, attribution debates, and governance instrumentation are reshaping contracts and negotiation dynamics. As AI industrializes, disciplined data economics — grounded in architectural clarity and life-cycle oversight — will become a defining competitive capability.

“AI is forcing structural renegotiation of data durability, value attribution, architectural control, and risk allocation across the AI life cycle.” — Lynne Schneider, research director, Data Collaboration and Monetization, IDC

From Pipettes to Playbooks: How AI Agents Are Rewiring the Life Sciences Workforce

Nimita Limaye — Thu, 19 Mar 2026 04:00:00 GMT

This IDC Perspective examines the adoption of AI agents in the life sciences industry, the drivers, the challenges, the use cases, and factors critical to driving adoption and provides critical guidance to buyers.

"Yes, the adoption of AI agents in the life sciences industry will continue to scale exponentially. Business process reengineering and workforce transformation will be critical to scale adoption and drive ROI. Determining accountability for both success and failure in a hybrid workplace will be challenging. HR will play a critical role in addressing concerns regarding both job insecurity and self-worth as AI agents gradually move up the value chain. HR and IT will have to work hand in hand to architect the hybrid workforce of the future," said Dr. Nimita Limaye, research VP, Life Sciences R&D Strategy and Technology at IDC.

IDC Future Enterprise Award: Special Award for Smart Cities- Best in Citizen Wellbeing Award Winner Profile 2025

Ravikant Sharma, Manoj Vallikkat — Thu, 19 Mar 2026 04:00:00 GMT

Taoyuan City built precision lung cancer screening model by combining online LDCT enrolment, AI assisted image interpretation, and mobile CT services. The program expands risk criteria beyond smokers and integrates cross department data triage to improve access and accuracy. Results from 2023 to 2025 show 92 percent early-stage detection, faster screening, high satisfaction, and strong economic and health system.

"Taoyuan City's lung cancer screening initiative demonstrates how smart city investments translate into measurable citizen wellbeing when risk intelligence, AI diagnostics, and outreach delivery are orchestrated as a unified public health service. By moving beyond smoker-centric criteria to population-specific risk stratification and mobile access, the program closes screening gaps, improves early detection at scale, and embeds equity into preventive healthcare design. It sets a strong benchmark for data-driven, inclusive health governance in smart cities," says Ravikant Sharma, Research Director, IDC Public Sector.

"As we enter into a year where enterprises shift from initiating their AI pivot and start to make investments in more advanced agentic AI systems, early disease detection takes new dimension and possibilities. Taoyuan's program stands out as it brings together digital enrollment, AI assisted interpretation, and mobile service delivery toward early detection at city scale. As AI evolves toward more autonomous agentic systems, this case reinforces a core requirement for public sector success. Focus should be on strong data governance, model oversight, and clear human accountability to scale impact," says Manoj Vallikkat, Senior Research Manager, IDC Public Sector.

IDC PlanScape: A Human Skills Framework for Agentic AI

Gina Smith, PhD — Thu, 19 Mar 2026 04:00:00 GMT

This IDC study sets out to break down generic skill categories into tangible, trainable capabilities. These include the ability to organize messy business questions into agent-sized tasks, explaining AI-assisted decisions to skeptical stakeholders, spotting when output needs human review and knowing when to override recommendations.

The framework's emphasis is on decision-making, collaboration, experimentation, and ethical judgment, with each tied to concrete subskills organizations can assess, teach, and track.

With the dedicated human skills framework described in this document, organizations can move beyond guesswork and take informed, strategic action.

"What enterprises need now is a plan," says Gina Smith, research director, IT Skills for Digital Business, IDC. "Organizations that have a strategy for assessing and mobilizing human skills alongside technical skills will be best positioned for AI transformation."

Inference at the Edge: The Next Frontier

Ghassan Abdo — Thu, 19 Mar 2026 04:00:00 GMT

Inference at the edge refers to executing AI model predictions locally on devices such as sensors, cameras, industrial systems, vehicles, or on-premises gateways rather than in centralized cloud datacenters. Although model training typically remains cloud-based because of its computational intensity, inference is increasingly deployed at the edge to enable real-time decision-making, reduce bandwidth consumption, enhance privacy, and ensure operational resilience in environments with limited connectivity. This shift is driven by use cases across manufacturing, retail, healthcare, telecommunications, energy, Smart Cities, and automotive sectors, where millisecond matters and data sovereignty or cost considerations make local processing more efficient and practical. Achieving this requires model optimization techniques (e.g., quantization and pruning), lightweight runtimes, AI-optimized silicon, secure device management, and integrated edge-to-cloud orchestration.

The document also highlights that edge inference represents a broader architectural transition from centralized AI to distributed intelligence, supporting Industry 4.0, 5G-enabled services, and digital transformation initiatives. However, organizations must address challenges such as hardware heterogeneity, limited compute and power resources, security risks, and the large-scale life-cycle management of distributed devices. A survey of providers — including Akamai, Cloudflare, AWS, Lumen, Tencent, and Telefónica — shows varied strategies, ranging from serverless AI platforms and global edge networks to infrastructure-led bare metal offerings and telecom-based distributed edge architectures. Collectively, these approaches reflect an evolving ecosystem focused on delivering low-latency, secure, and scalable AI inference closer to where data is generated.

"Inference at the edge represents a pivotal shift in enterprise AI strategy, moving intelligence from centralized clouds to the point of data creation. Organizations that successfully deploy edge inference will unlock real-time decision-making, reduce operational costs, and strengthen data sovereignty while enabling new Industry 4.0 and 5G-driven use cases. However, success will depend on integrating optimized models, secure device management, and scalable edge-to-cloud orchestration to manage distributed complexity and deliver measurable business outcomes," says Ghassan Abdo, research VP, Worldwide Telecom.

MWC26 Operations and Monetization Updates: AI, Agentic

Chris Silberberg — Thu, 19 Mar 2026 04:00:00 GMT

Mobile World Congress 2026 highlighted the telecom industry's shift from technology experimentation to execution, with agentic AI, data quality, and telecom ontologies as central themes. While autonomous networks are advancing, data fragmentation, trust in AI-driven autonomy, and workforce transformation remain significant hurdles.

Modernizing Payroll: Adaptability, Compliance, and AI

Zachary Chertok — Thu, 19 Mar 2026 04:00:00 GMT

This IDC Perspective highlights practices for modernizing payroll. Payroll does not happen in a silo, even if it is currently managed within one for several organizations. Integrating payroll as a process within broader HCM is vital as payroll dependencies and requirements often bookend critical workforce data, insights, and decisions. Modernizing payroll relies on tighter and more natively intuitive integrations between systems of pay and systems of management. At the top of the list for modernization, payroll leaders need to examine how AI can be used to mirror employee records and associated pay valuation changes directly and compliantly into the payroll environment, all while maintaining guardrails for rules-based adherence toward accurate and on time pay.

"Payroll leaders are looking to guarantee pay accuracy and rigid compliance as they always have while embracing AI to help them manage more frequently changing characteristics of employee position-based pay," says Zachary Chertok, senior research manager for HCM Applications and Agents at IDC. "Looking for deterministic guarantees using probabilistic systems links is challenging payroll teams to rethink what they need to own and manage as the train to adaptable operationalization is pulling away from the station without them onboard. Workflow avenues for AI can benefit payroll leaders' abilities to keep up while allowing them to be unyielding for accuracy and compliance."