AI isn’t just software—it’s an operating-model redesign. Lori Bieda argues that talent, trust, and governance will determine who truly transforms.
In corporate boardrooms across North America, artificial intelligence is still introduced with the language of software: deployment, models, vendors, and platforms. Budgets are approved. Slide decks circulate. Pilots are announced.
But from where Lori Bieda sits, this framing misses the point.
Bieda is an AI, data and digital transformation executive who has led enterprise analytics at institutions including BMO, SAS, CIBC and TD. She has spent three decades building data-driven organizations in some of the world’s most regulated industries. She now advises governments and boards while leading ElleExcel Women’s Circle, a global platform focused on women’s career progression.
Her conclusion is stark: AI transformation is not a technology rollout. It is an operating model redesign disguised as one.
“AI is a profound change management initiative that happens to be enabled by technology,” Bieda told me. “It changes how we organize teams, how we make decisions, how work flows, and ultimately what work even exists. That’s structural change, not software change.”
Technology, she says, is merely the spark. Leadership and operating model are the fuel.
The real question facing executives is not whether they can build models. It is whether they can reshape how the company runs with AI at the center.
Sailboats Before Tankers
There is a growing narrative that corporate AI efforts are faltering — that pilots stall, budgets shrink, and proofs of concept gather dust. Bieda rejects the word “failure.”
“We are not failing — we are testing,” she said. “And testing always comes with learning.”
What she sees inside large enterprises is disciplined experimentation. Contained AI agents supporting workflows. Narrow automation is replacing repetitive tasks. Human-in-the-loop systems absorb workload rather than redefining the enterprise overnight.
“Right now, AI is mostly absorbing workload, not redefining companies,” she said. “Transformation follows traction.”
The metaphor she favors is nautical. Companies are launching sailboats, not tankers — contained use cases designed to study the currents before committing capital at scale. Governance is stress-tested. Data weaknesses surface. Human resistance becomes visible. Boundaries flex.
This is not recklessness. It is prudent.
Industry data supports the caution. A significant portion of generative AI initiatives remain in pilot stages. Broad, cross-functional transformation is still rare. For Bieda, that is evidence of responsibility, not retreat.
“Most enterprises are approaching AI through a practical, value-first lens,” she said. “Leaders are looking for beachheads — small wins that free up budget, reorder the cost model and build trust before scaling bigger changes.”
The Plumbing Problem
If there is frustration within organizations, it is often directed at data governance—the unglamorous plumbing beneath AI’s sleek interface.
Why chase innovation before fixing the pipes?
“AI hasn’t created the plumbing gaps — it’s exposed them faster and more visibly,” Bieda said.
Introduce AI into a workflow, and dormant issues surface immediately: inconsistent definitions, weak lineage, fragmented systems, and thin metadata. What once took quarters to diagnose now appears in days.
For data leaders, she argues, this is a once-in-a-generation opportunity.
“If organizations want the payoff from AI, they have to invest in the fabric of their data,” she said. Without enrichment, lineage, and guardrails, fully agentic systems remain fantasy. At best, companies automate inefficiency. At worst, they scale bias and error.
Governance, in her framing, is not friction. It is a monetization infrastructure.
“AI turned the lights on,” she said. “For data leaders, this is the moment to fund the foundations that make monetization possible.”
Advice or Manipulation?
Nowhere is the ethical line more delicate than in financial services, where Bieda spent years embedding AI into retail and business banking.
Personalization, she argues, is not inherently manipulative. At its best, it is advisory.
“When personalization is done well, it creates a symmetric value exchange,” she said. “Personalization should feel like advice, not persuasion.”
The line is crossed when nudges prioritize profit over outcomes, when transparency disappears, or when data is used for purposes customers did not intend.
Yet the boundary itself is shifting. As consumers grow more comfortable sharing data, their expectations rise in parallel. They want companies to “connect the dots.” Trust expands when value is obvious — and contracts when it is not.
The imperative is dynamic equilibrium: continuous temperature checks with customers as norms evolve.
AI and the Gender Gap
Artificial intelligence is often described as a democratizer. But technology learns from history, and history contains bias.
Before founding ElleExcel, Bieda served as Chief Data, Analytics, and AI Officer at BMO for the North American Retail and Business Bank, overseeing more than 500 analysts across geographies. She has also spent 25 years as a woman in STEM.
“AI learns from history,” she said. “And history contains bias.”
From medical research centered on male bodies to financial models built on male income trajectories, bias is embedded in the data exhaust of past decades. Unchecked, AI scales those inequities faster.
“AI will scale whatever system you feed it,” she said. “We’re choosing to scale opportunity.”
Through ElleExcel, Bieda is building tools that use data and AI to strengthen clarity, confidence and community for women navigating mid-career inflection points. The goal is not symbolic empowerment but structural leverage — putting intelligence directly in the hands of those historically underrepresented in leadership.
AI, she insists, can either automate inequity or amplify access.
Talent as the Accelerator
Five years from now, what uncomfortable truth will companies wish they had confronted earlier: talent, trust, or technological debt?
Bieda does not hesitate.
“AI only moves at the speed people move inside an organization,” she said. “Technology doesn’t create value on its own.”
She recounts deploying an AI agent designed to help thousands of frontline employees navigate thousands of policies daily. The tool worked. Adoption lagged.
Some employees assumed it could not handle complex questions. They defaulted to managers. Only when she sat beside an employee and asked her to pose a real query did trust shift. The agent responded accurately, with full context and supporting documentation.
“You could see the trust click,” she said.
The barrier was not the algorithm. It was perception and behavior.
If people do not use the system, the system does not improve. And many still approach AI as if it were a search engine rather than a collaborative reasoning tool that requires structured prompts and critical thinking.
“AI doesn’t replace people,” Bieda said. “It raises the premium on people who know how to use it.”
Trust, she adds, is the guardrail. Technology is the engine. Talent is the accelerator.
The Real Exam
The AI moment is not a race to deploy the most models or announce the most pilots. It is a test of leadership maturity.
Can executives redesign workflows? Can they invest in invisible infrastructure? Can they educate their organizations to ask better questions? Can they embed ethics before scale? Can they build trust before automation outpaces comprehension?
AI may be enabled by code. But its success is determined by people.
In the end, the companies that thrive will not be those that merely installed the technology. They will be those who reimagined how they operate — and who is empowered to operate within it.
