PERSPECTIVE
Want more value out of AI? Redesign your operating model with security built in.
5-MINUTE READ
November 17, 2026
PERSPECTIVE
5-MINUTE READ
November 17, 2026
Companies are spending more on AI. Yet just a quarter of executives report widespread and sustained value from their investments.1 Gaps in data, the wrong tools and uneven adoption all play a part. But the deeper issue is the operating model. It needs to shift to design for how people work with AI and to address heightened threats and vulnerabilities that arise as AI changes who makes decisions and where.
Cybersecurity must be embedded into that shift to accelerate AI value. When clear ownership, guardrails and trust are built into how people and AI operate, teams can innovate confidently at speed and scale. Cybersecurity protects the things that make AI valuable such as proprietary context, expertise and decision history.
Leading companies show what’s possible. Among those with the most mature security postures, 73% ensure board-level accountability and embed security into strategic decision-making. They also see 1.6x higher returns on AI and 1.7x bigger reduction in technical debt than the least mature organizations.2
Other companies can follow their lead and realize value faster with three moves:
Organizations are already struggling with a shortage of cybersecurity talent,3 and AI places more demands that widen the gap. It raises the bar for skills needed while simultaneously expanding the threat surface, and therefore cybersecurity’s remit. Cyber professionals need AI skills, AI-enabled delivery capability and an understanding of AI risk today. These skills sit on top of the specialty expertise the role demands already.
Meanwhile, employees across functions can configure AI systems, connect proprietary data to foundation models and deploy agents with little IT support. This democratization can accelerate innovation. But it also moves important security decisions outside of traditional governance processes.
This raises a question every member of the C-Suite must answer: Can the current workforce meet these needs by making the transition? Some teams may be able to bridge gaps through upskilling. Others may need different reskilling approaches or workforce models. The answer decides whether organizations can scale AI securely or whether talent becomes the constraint that limits the value they get from it.
Most cybersecurity functions were not designed for today’s AI-enabled threat landscape. Attackers can now execute with lightning speed, with average breakout times falling to 29 minutes, and the fastest on record was 27 seconds.4 This means an organization can suffer impact before it even knows it has been breached.
Cybersecurity strains as employees and agents build, decide and act with AI, and threat actors use the same technology to speed up attacks. When security decisions become more than the central function can handle, a false choice emerges. If every AI use case waits for central approval, innovation stalls. But risk accumulates when teams route around security, exposing the very assets which make AI valuable: proprietary context, expertise, decision history and organizational judgment.
This risk is two-fold: attackers can steal these assets, putting companies and people at risk, or corrupt entire workflows. Think deliveries routed to the wrong location or supply chain temperature controls switched off.
As it is, 52% of Chief Information Security Officers say their scope is no longer fully manageable.5 And 53% of employees say they do not know who is accountable when AI makes an error.6 These are early indicators of cracks in the model that can eventually spread throughout the enterprise.
of Chief Information Security Officers (CISOs) say their scope is no longer fully manageable.5
of employees say they do not know who is accountable when AI makes an error.6
Competing in the AI-era requires the enterprise operating model to evolve. This comprises leadership alignment, workforce planning and skills development, culture change and organizational transformation for sustainable AI adoption.
In a recent report,7 we identified five shifts that are crucial to this evolved model:
Each shift distributes authority closer to the work, accelerates the pace of execution and expands the circle of people building with AI. To remain resilient, business leaders need to mull over new questions: Who owns accountability when decisions move into integrated value streams? How do you divide decision authority among humans and AI? How do agents get governed when anyone can deploy them? What protects the proprietary assets that define competitive advantage, when underlying AI models are available to every competitor?
To answer these questions, security must be built into the shifts from the outset; not as a constraint on what companies can do but as a condition for democratized innovation. This separates companies building durable AI advantage from those seeing their AI investment outpace value, or worse, watching trust, reputation and revenue erode.
Three moves enable these shifts with cybersecurity as a value driver.
Securing AI-enabled business outcome is a shared C-Suite responsibility. The Chief Executive Officer leads with business ambition, holding the entire C-Suite accountable for both AI value and cyber resilience. The Chief Operation Officer orchestrates, turning that mandate into operations. For instance, a 2025 report showed distributed leadership; 57% of respondents said their head of cybersecurity reports to the head of IT such as the Chief Technology Officer or Chief Information Officer, while one-third said the cybersecurity leads report to another C-Suite executive.8 The CISO remains the “captain,” or the enterprise authority on cyber risk and cyber mitigation, supported by every business leader.
Across the C-Suite, investment, technology, risk and execution decisions must reinforce one another. The Chief Financial Officer ensures every AI investment is supported by the cybersecurity capabilities needed to protect its value, without which even the most promising initiatives can fall short of expected returns.
Each leader should own the cybersecurity and resilience of the AI, data and agents used to deliver their business outcomes. These responsibilities belong on executive scorecards alongside revenue, cost and growth. Cybersecurity is then measured the same way the business measured itself: As part of the cost and resilience of the outcomes delivered by a value stream.
Such distributed accountability requires strong protection, detection, response and recovery during a cyber event, with the CISO and COO leading from the front. Shared responsibility doesn’t mean blurred ownership: Our research shows 42% of companies spread accountability for cyber resilience across multiple C-Suite leaders, yet 46% lack a single, clearly accountable owner for continuity when disruption hits.9
Accenture’s own incident response is overseen by a cross-functional team that includes CISO, our General Counsel and Chief Marketing Officer. Combined with a board Audit Committee that oversees cyber risk and an extensive governance network that connects markets, services and functions, we’ve cultivated the shared accountability to support stakeholder trust.10
When investment, operations and cybersecurity decisions come together, teams can innovate within clear limits, and business leaders remain accountable for sustaining that value even through disruption.
Many AI programs change the technology but leave rest of the organization intact. That is why adoption outruns value and security becomes the last signature on decisions it never shaped. Sustainable and even adoption at scale combines workforce planning and skills development with operating model redesign, business process change, culture change and leadership alignment.
For this, leaders must start with three questions: What capabilities do we need? What capabilities do we have? How do we close the gap? Answering them means mapping existing skills, identifying where AI creates new cybersecurity requirements, and deciding which and how talent can be upskilled, reskilled or sourced. Recruitment, skilling and workforce models will vary across regions, depending on the nature of talent supply and regulations.
While AI is exacerbating cybersecurity’s talent constraints, it also offers organizations a mechanism to overcome them. The question is how leaders redesign the workforce and operating model to capture that opportunity with a culture of security built around it.
In the absence of such culture, employees and agents often choose the path of least resistance, using the fastest tool or the ungoverned model, to get the job done. Over time, those choices normalize insecure behavior. AI can amplify the cost of such behavior, especially during a breach because the proprietary knowledge and judgment that set an organization apart now flow through the same systems.
Meanwhile, attackers are using AI to find vulnerabilities faster,11 scale social engineering and exploit the gaps that result from rapid adoption. These gaps multiply when agents lack named owners and a defined remit, and employees deploy AI without adequate skills or guardrails. Business leaders can create a security-first culture by establishing incentives and safeguards and equipping employees to identify risks and use AI safely, so the secure path is also the easiest one.
A global financial institution embarked on a risk transformation journey with its 50,000+ employees at the center. We helped strengthen its risk culture, bring clarity to the new operating model and ways of working, and reinforce the lines of defense. Ongoing learning and engagement will keep pace with an evolving risk landscape.
When everyone has access to similar AI, advantage comes from assets such as the distinctive knowledge and creativity that a company brings to it. AI can scale that advantage. It can also expose or dilute it. A redesigned enterprise operating model can protect these assets.
Today’s new crown jewels in business are proprietary context, accumulated expertise, decision history and the organization judgment created through applying AI. These distinguish a business and fuel its ability to adapt. Cybersecurity, built into the operating model, can protect them alongside systems and data. As they flow through prompts, models and retrieval systems, controls must follow the assets.
Decision history is a proprietary learning asset that grows stronger with every decision made,12 creating an advantage over competitors running the same AI models. It records the context, AI recommendations, human overrides and outcomes behind each choice. That record is a differentiator and a key driver of value from AI. Protect it from the very first decision loop: assign an owner, track its origin, restrict access, set retention rules and specify which agents and models may read or change it.
Mastercard, for example, is using predictive AI technology that identifies links between potential card or merchant compromises. As new data comes in, links are either created or dropped to improve pattern recognition. The technology doubled detection rates and sped up at-risk merchant identification by 300%.13, 14
Cybersecurity must follow the work, from strategy and investment to AI inputs and outputs. It must be therefore embedded throughout the operating model. Clear boundaries free teams to innovate, resilience keeps critical operations running through disruption and adaptability lets the business change course with market conditions. Together, they protect today’s business and preserve its capacity to compete tomorrow.
Leadership now has specific roles to play and questions to answer:
Answering these questions will help redesign the enterprise for lasting advantage without surrendering control of what makes a business distinct.
Thank you to our contributors, Shelby Flora and Kent McMillan
¹ Accenture Pulse of Change, Oct. 19, 2026
² Accenture State of Cybersecurity Resilience 2025, June 25, 2025.
³ Reinventing the Cyber Workforce | Accenture, June 2, 2026
⁴ 2026 CrowdStrike Global Threat report, February 24, 2026.
⁵ 2026 State of CISO Benchmark Report, IANS and Artico Search, January 15, 2026.
⁶ The CHRO as architect of enterprise value, Accenture, August 25, 2026.
⁷ Operating Model in the Age of AI: Five Shifts | Accenture, October 6, 2026
⁸ 2025 State of Cybersecurity Report: An In-house Perspective, ACC Foundation, March 24, 2025.
⁹ Redefining Cyber Resilience, Accenture, June 30, 2026. N = 1,000. C-Suite leaders
¹⁰ Accenture Annual Report: Fiscal Year 2026, Form 10-K, Item 1C, October 16, 2026.
¹¹ 2026 CrowdStrike Global Threat report, February 24, 2026.
¹² Future of Work: Work That Matters | Accenture, 2026
¹³ Mastercard accelerates card fraud detection with generative AI technology, Mastercard, May 22, 2024.
¹⁴ Cuttler, Elyse, Inside the algorithm: How gen AI and graph technology are cracking down on card sharks, Mastercard, July 18, 2024.