Kyndryl’s AI modernization surges—but security woes loom

Kyndryl's Agentic Modernization platform, powered by the Kyndryl Agentic AI Framework and underpinned by Kyndryl Bridge, has demonstrated transformative impacts by reducing IT incidents by up to 50% across more than 1,400 enterprises. Generating over 16 million AI insights monthly, the platform codifies Kyndryl's mission-critical engineering expertise into pre-defined, AI-orchestrated workflows that shift modernization from episodic, labor-intensive projects to continuous, predictable operating disciplines, enabling faster execution and significant operational efficiencies such as halving data center exit timelines.

At Suryoday Bank, Kyndryl’s Agentic Modernization platform has been instrumental in integrating AI-powered workflows that enhance compliance, operational efficiency, and customer service by automating tasks like digitizing account-opening forms and providing voice-enabled mobile banking support. Beyond improving productivity and reducing response times, the platform supports critical regulatory functions such as legal enforcement responses and MSME loan underwriting, facilitated by Kyndryl’s consult-led approach and a dedicated Agentic AI Center of Excellence that scales AI agents securely across complex banking operations.

Kyndryl’s technical framework, particularly the Kyndryl Bridge, provides robust deployment, orchestration, and governance capabilities for AI agents and workflows across hybrid IT estates, ensuring enterprise guardrails and cost visibility throughout AI adoption. This infrastructure readiness, combined with flexible model usage favoring open source but accommodating proprietary customer models, enables scalable agentic AI integration that aligns with ITIL best practices and maturity models, helping clients progress toward ISO certification through incremental automation of IT service management processes.

While broad horizontal adoption of agentic AI remains nascent, Kyndryl is actively driving vertical use cases in sectors like healthcare and customer onboarding, and is pioneering automation in IT service management through its agentification of a four-year-old AIOps platform. By converting well-defined ITIL processes into autonomous or semi-autonomous AI agents tailored to customer preferences, Kyndryl is creating a 'modernization dividend' that could reduce IT service management costs by up to 90%, freeing resources for further enterprise modernization.

Enterprise AI transformation demands a fundamental rewiring of organizations to embed intelligence into every workflow, decision, and customer interaction, moving well beyond isolated pilots. Dr. Sanjeev Rastogi underscores that the next competitive advantage belongs to businesses that treat data as a strategic asset, build AI literacy across all levels, and establish strong governance frameworks emphasizing data protection, human oversight, and transparency. This holistic approach fosters a culture of experimentation and continuous reinvention, ensuring AI investments drive measurable outcomes such as revenue growth, operational agility, and trust rather than mere cost savings.

Scaling agentic AI within enterprises hinges on meticulous workflow redesign and human-in-the-loop governance to balance automation with necessary oversight. Companies like PromptCare and DXC Technology exemplify this by mapping end-to-end processes before automation and automating routine tasks such as SOC alert triage, thereby freeing human experts for complex decisions. This approach requires clear organizational design decisions about which actions can be automated, which require escalation, and which must remain human-led to avoid bottlenecks, while fostering cultural acceptance through transparent communication about AI’s role.

The evolution of AI operating models elevates developers from mere coders to architects and orchestrators who design multi-agent workflows, enforce governance, and ensure responsible AI implementation. As Ganesan Karuppanaicker of Birlasoft notes, successful scaling depends on foundational readiness—including high-quality data, integrated cloud platforms, and AI-ready talent—and embedding governance during execution through policy-as-code and real-time oversight. This shift transforms AI adoption from a technology upgrade into a comprehensive business transformation program with leadership accountable for linking AI investments to productivity, customer experience, and growth outcomes.

Operationalizing AI at scale requires composable, adaptive architectures and robust governance platforms like AccuKnox’s AgentZ, which consolidate agent management, sandboxed execution, zero-trust controls, and auditability to mitigate risks and enable enterprise-wide deployment. This infrastructure supports multi-team organizational models with role-based access and dynamic policies, facilitating secure, scalable AI integration that moves enterprises from fragmented pilots to coherent, outcome-driven programs. However, challenges remain in underinvested areas such as AI agent evaluation and testing, underscoring the need for continuous refinement to realize AI’s full business value.

Kyndryl and AWS have significantly accelerated cloud migration processes by leveraging a sophisticated multi-agent AI orchestration framework on Amazon Bedrock AgentCore. This framework automates critical migration phases—from discovery and infrastructure as code generation to governance and post-migration operations—reducing infrastructure as code development time from weeks to mere minutes across portfolios exceeding 300 applications. Complemented by AWS managed services like Database Migration Service with generative AI-assisted schema conversion and AWS Transform for legacy code modernization, this collaboration streamlines complex migrations while ensuring security and operational resilience.

The expanded Kyndryl-AWS alliance underscores a strategic push to embed agentic AI deeply into enterprise modernization efforts, exemplified by their co-development of AI modernization blueprints and Kyndryl’s new status as an Anthropic Authorized Reseller for Amazon Bedrock. This partnership not only enhances AI adoption at scale but also reflects Kyndryl’s growing expertise, as evidenced by earning three new AWS Competencies in Mainframe Modernization, AI, and Digital Sovereignty. Real-world client successes, such as Alpitour World’s cloud-native mainframe redevelopment yielding 25-30% cost reductions and AI-powered sales automation, highlight the tangible business impact of this collaboration.

Beyond public cloud modernization, Kyndryl’s expanded partnership with Broadcom focuses on advancing AI-driven private cloud modernization by investing heavily in training thousands of certified consultants skilled in VMware Cloud Foundation 9.1. This initiative addresses a critical talent shortage in deep VMware expertise, which industry analysts like Mike Leone identify as a major bottleneck for hybrid infrastructure projects. By enabling private clouds to emulate public cloud capabilities through automation, policy as code, container support, and AI governance, Kyndryl and Broadcom aim to simplify the complexity of hybrid AI infrastructure and unlock real business value for enterprises.

Kyndryl's Q1 financial results underscore persistent profitability challenges, with a net loss of $55 million on $3.618 billion in sales, signaling ongoing revenue softness that weighs heavily on investor confidence. Despite a substantial share repurchase of over 19 million shares totaling $462.67 million aimed at bolstering shareholder value, concerns about deal slippage and fragile revenue streams continue to cloud the outlook, suggesting that capital return efforts may not fully offset underlying operational risks.

Investor sentiment has been further shaken by a recent cybersecurity incident involving compromised credentials, which threatens to erode Kyndryl's reputation as a 'gold standard' in IT services and raises questions about its pricing power and risk management. This breach, while not altering fundamental business metrics, casts a shadow over the company's security posture and may prompt analysts to reassess previously optimistic earnings forecasts, including projections of $569 million in earnings and an 11% upside to a $14.10 fair value target.