The New AI Digital DividesTM
Why Access Is No Longer Enough – Why Senior Leaders Must Build Capability Before the Gap Becomes Permanent
A Collaboration Evangelist Underwood Partners White Paper
Craig Underwood | July 2026
I thought I understood the emerging divide around artificial intelligence. On one side were people and organizations using tools such as ChatGPT, Gemini, and Claude; on the other were people who were not using them, could not use them, or were not permitted to use them.
Then I began building agents.
At first, I used AI much as many professionals do: to research questions, improve writing, brainstorm, and make sense of complicated information. With training and experience, I learned to prompt more deliberately: provide context, define the task, request a format, ask the model to critique itself, compare options, and refine the output. That alone substantially changed the quality of the work.
But the bigger change occurred when I began building custom agents that could perform multi-step research and analysis workflows. Suddenly, I was not simply asking AI for a better answer. I was delegating meaningful parts of a workstream. The challenge began to shift from producing output to supervising, evaluating, synthesizing, and acting on far more output than I could previously create myself.
That experience caused me to rethink the phrase digital divide.
The original digital divide concerned access: who had a computer, who had broadband, who could participate in the internet economy. Access to AI remains important. But it is no longer the only divide, or perhaps even the most consequential one for organizations. A widening set of capability divides now separates people and enterprises that cannot or will not use AI from those that use it casually, those that use it skillfully, those that build agents to execute work, and those that coordinate suites of agents across workflows.
My hypothesis is simple:
The next competitive, workforce, and social-equity challenge is not one AI divide. It is a ladder of AI capability divides – and the distance between the rungs is beginning to matter.
The research published over the past two years strongly supports this hypothesis, with one essential qualification: these five levels are not neat, mutually exclusive demographic categories. Organizations may occupy several at once. A hospital might prohibit public tools for patient information while deploying secure AI in administrative work. A firm may have Level 2 employees and Level 5 technology pilots at the same time. This is not a scorecard for judging people. It is a leadership framework for recognizing capability gaps before they harden into inequality and competitive disadvantage.
Figure 1. The Five-Level AI Capability Ladder.
The five divides
Level 1: Excluded, restricted, or refusing AI
This first category is more complex than “anti-AI.” Some people actively reject AI for philosophical, labor, environmental, artistic, or ethical reasons. Others want to learn but lack access, confidence, or opportunity. Still others work in organizations that appropriately restrict public AI tools because of privacy, confidentiality, regulated data, intellectual property, or safety concerns.
The distinction matters. A responsible restriction on feeding sensitive customer, patient, client, or proprietary information into a public model is not backward-looking. It is governance. But an organization that stops at prohibition – without providing approved alternatives, clear policies, and training in the responsible and productive use of AI – may unintentionally place employees at a capability disadvantage while also driving “shadow AI” use underground.
That is not a theoretical risk. The University of Melbourne and KPMG’s 2025 global study found that 70% of employees intentionally using AI at work accessed free, publicly available AI tools, while only two in five reported that their organization had a policy or guidance on generative AI use. Nearly half reported uploading sensitive company information or copyrighted material into public AI tools; strikingly, the behavior was most common among employees reporting that their organization had banned generative AI. Cisco’s 2026 Data and Privacy Benchmark Study, based on a September 2025 survey of more than 5,200 IT, technology and security professionals with data privacy responsibilities across 12 markets, adds an institutional view: 90% said AI had broadened the scope of their privacy programs, but only 12% described their AI governance committees as mature and proactive. The policy challenge is therefore not whether leaders should ignore risk. It is whether they can protect data while still enabling safe learning and useful experimentation.[1]
Figure 2. AI use is outrunning governance.
There is also a growing institutional access gap. The OECD reported in 2025 that AI adoption divides are forming along existing fault lines among places, sectors, and firms. Its analysis found that AI diffusion in 2023-24 was being driven more by leaders “escaping the pack” than by laggards catching up.[2] Put differently: as AI adoption accelerates, previous disadvantages in capital, skills, digital infrastructure, and organizational capacity may be magnified rather than erased.
Level 2: AI-assisted tasks – the “better Google and Grammarly” stage
This level captures the vast middle of current AI adoption: employees use AI to locate information, consolidate documents, generate ideas, summarize meetings, rewrite emails, or improve drafts. This is valuable. It can save time, reduce friction, and introduce people to a powerful technology.
But it is still mostly individual, task-level augmentation rather than a redesign of work.
Gallup’s April 2026 study of 23,717 employed U.S. adults found that half use AI in their role at least a few times a year, while 28% use it at least a few times a week and 13% use it daily.[3] A prior Gallup analysis found that common uses were precisely the kinds of activities associated with Level 2: consolidating information, generating ideas, learning new things, and using chatbots or writing/editing tools.[4]
This is a genuine productivity opportunity. It is also a potential trap for leaders who confuse widespread experimentation with organizational transformation. In the same 2026 Gallup study, 65% of employees in organizations implementing AI said it had improved their productivity and efficiency; yet only about one in ten strongly agreed that AI had transformed how work gets done in their organization.[5]
Figure 3. AI adoption is broad; deep use is not.
The distinction is crucial. If employees become slightly faster at emails, summaries, and brainstorming while the organization continues operating through the same slow handoffs, redundant meetings, outdated processes, disconnected systems, and approval bottlenecks, the organization has captured convenience – not transformation.
Level 3: AI-literate knowledge work
At Level 3, AI is no longer simply a search box or writing assistant. People have learned enough to work with it deliberately. They can frame a problem, supply context and source material, specify the desired deliverable, challenge weak outputs, verify claims, ask for alternatives, iterate toward a stronger result, and understand when not to rely on the model.
I use the phrase AI-literate knowledge work intentionally. Prompt engineering matters, but this level is broader than clever prompts. It includes judgment, verification, data protection, task selection, critical thinking, and the discipline to keep humans accountable for important decisions.
Research suggests this divide is already material. Boston Consulting Group’s 2025 global AI at Work survey found that only 36% of employees felt properly trained in the skills needed for AI transformation and only 25% of frontline employees reported receiving sufficient support from leadership on how and when to use AI. Yet regular AI use was closely associated with training intensity: 18% among employees receiving no training, 63% among those receiving one to five hours, 82% among those receiving five to ten hours, and 89% among those receiving more than ten hours.[6]
Figure 4. Training and leadership support lag behind adoption.
This is where the AI divide becomes a leadership issue. When executives offer a tool but do not provide training, permission boundaries, safe data practices, coaching, example workflows of AI prompts/ instructions that worked – and extremely important, those that didn’t work – or time to practice, they create an environment in which the confident and self-directed advance while others remain behind. In effect, they outsource capability building to personal curiosity, spare time, and professional privilege.
The stakes are economic as well. PwC’s 2025 Global AI Jobs Barometer, drawing on nearly one billion job advertisements across six continents, found a 56% wage premium for jobs requiring AI skills compared with similar jobs without those skills. It also found that skills are changing 66% faster in AI-exposed jobs than in other jobs, and that industries more exposed to AI experienced three times greater growth in revenue per employee.[7] This does not prove that prompting alone produces higher wages or productivity. It does strongly suggest that the capacity to work effectively with AI is becoming economically valuable – and rapidly so.
Importantly, AI literacy can sometimes narrow existing skill gaps rather than worsen them. A peer-reviewed 2025 study in The Quarterly Journal of Economics examined 5,172 customer support agents using an AI conversational assistant. Access to AI increased productivity by 15% on average, with the largest gains accruing to less experienced and less skilled workers; novice workers with AI reached performance levels comparable to far more experienced workers without it.[8] This is a hopeful finding: appropriately deployed AI can democratize expertise. But it also underlines the moral and organizational consequence of denying some workers the training and tools that allow them to benefit.
Level 4: Agent-enabled workflows
At Level 4, the user is not simply conversing with AI. The user creates or deploys an agent that can conduct a multi-step task: search defined sources, analyze data, prepare a first draft, compare options, produce a report, monitor a condition, or support a repeatable operational workflow.
This is the divide I only fully appreciated after beginning to build custom agents myself. There is a dramatic difference between asking an AI system, “What companies might sponsor this nonprofit event?” and designing an agentic workflow that methodically reviews industry lists, identifies relevant companies, researches corporate giving, checks foundation tax filings, ranks prospects, records findings, and produces a briefing for human review. The human remains accountable. But the scale and speed of possible work changes materially.
Early organizational examples illustrate this shift.
Moderna: After providing employees access to ChatGPT Enterprise, Moderna reported that within two months it had 750 custom GPTs across the company; 40% of weekly active users had created a GPT; and the average user conducted 120 ChatGPT Enterprise conversations per week. One pilot, Dose ID, was designed to review and analyze clinical data and support clinical study teams’ dose-selection analysis while preserving human judgment and review.[9] Moderna’s own description emphasizes that embedding AI requires not just technology but an “AI-focused culture” and on-the-job training.[10]
Morgan Stanley: In 2024, Morgan Stanley Research announced AskResearchGPT, an AI-powered capability intended to help client-facing teams find and synthesize insights from the firm’s research. The company described it as part of a growing suite that included AI @ Morgan Stanley Assistant and AI @ Morgan Stanley Debrief for wealth-management advisors and staff.[11] This is not an employee polishing an email. It is a firm creating controlled, domain-grounded AI capability embedded in professional work.
Small and midsized organizations: The opportunity is not reserved for global enterprises. In a representative survey of more than 5,000 small and medium-sized enterprises across seven OECD countries, 31% reported using generative AI. Among SME users, 65% said it had improved employee performance, 39% of those experiencing a skills gap said it helped compensate for that gap, and one-third reported reduced workload. Yet a third or fewer were taking measures to train staff, set internal guidance, or research legal and regulatory issues.[12] This combination – useful technology without commensurate capability building – is precisely how a new divide can widen.
Level 5: AI-orchestrated organizations
At the fifth level, organizations coordinate multiple agents, workflows, tools, data sources, and human decision makers. Humans set objectives, define permissions, supervise exceptions, evaluate results, protect values, and bear accountability. Agents may research, draft, schedule, classify, analyze, retrieve, monitor, escalate, or execute bounded tasks across a connected process.
This is still emerging. Leaders should be wary of hype and “agentwashing,” in which an ordinary chatbot is described as an autonomous agent. Yet the direction of travel is increasingly visible.
Microsoft’s 2025 Work Trend Index, based on survey data from 31,000 workers across 31 countries as well as labor-market and productivity signals, described an evolution from AI as assistant, to agents as “digital colleagues,” to humans directing agents that run whole workflows. It found that 81% of leaders expected agents to be moderately or extensively integrated into their AI
strategy within 12 to 18 months, and 46% said their companies were already using agents to fully automate workflows or processes.[13]
The report also reveals an agentic gap between leaders and employees. Sixty-seven percent of leaders reported familiarity with agents, compared with 40% of employees. Sixty-nine percent of leaders used AI regularly, compared with 45% of employees. And 36% of leaders expected to manage agents in their jobs, compared with 21% of employees. Most strikingly, 42% of leaders expected their teams to be building multi-agent systems to automate complex tasks within five years.[14]
Figure 5. The frontier is moving from use to orchestration.
Technology suppliers are now making this capability easier to deploy. Microsoft announced multi-agent orchestration in Copilot Studio in 2025, enabling agents to delegate tasks and share results across complex workflows.[15] Anthropic described how its own multi-agent research system uses a lead agent to coordinate parallel subagents, reporting that the multi-agent configuration outperformed a single-agent configuration by 90.2% on its internal research evaluation for breadth-first research tasks.[16] Gartner forecast in August 2025 that task-specific AI agents would be integrated into 40% of enterprise applications by the end of 2026, up from less than 5% in 2025; Gartner also explicitly warned against confusing AI assistants with agents.[17]
None of these findings means every organization should rush to automate every process. Agentic systems introduce new risks: security exposures, unreliable actions, poor source selection, errors at scale, loss of accountability, employee anxiety, and the possibility that leaders automate activity without improving outcomes. They require more governance, not less.
But they do reinforce the central hypothesis: organizations capable of designing, supervising, and safely coordinating agents may operate in a fundamentally different productivity environment from organizations still debating whether employees may use an AI writing assistant.
Why this is not merely a technology story
The new AI divides matter for at least three reasons.
1. They are a competitive and economic issue
The difference between Level 2 and Level 4 is not simply efficiency. It is leverage. A professional using AI to revise an email may save minutes. A professional managing a well-designed research agent may expand the number of opportunities evaluated, accelerate decisions, and produce capabilities that previously required a team or an outsourced engagement.
The evidence is not yet sufficient to calculate an exact productivity premium for each rung of this ladder. Indeed, Gallup’s research properly cautions that individual task gains have not yet translated into widespread transformation of organizational work. But that is precisely the point: organizations capable of redesigning workflows may separate themselves from organizations that never move beyond incidental personal usage.
2. They are a workforce development issue
A digital divide becomes socially consequential when the people most in need of capability are least likely to receive it. Frontline employees, workers in smaller organizations, professionals outside technology and finance, and communities with fewer learning resources risk becoming users of yesterday’s workflows while better-supported peers learn to direct AI-powered work.
I believe leaders have the responsibility – some would say the moral imperative – to help employees learn how to use AI responsibly and efficiently and to move toward higher-value work involving judgment, relationships, creativity, influence and accountability. This is especially important for nonprofit leaders of work-force development and college access/ success organizations, and those whose missions include preparing those they serve for jobs of the future. Imagine the different impression a young entry-level job or internship applicant would leave on a prospective employer when answering the question “Do you use AI?” between:
“[My training organization] has provided responsible AI literacy training and prompt engineering to all of its students.”
And
“I don’t use AI much at all.” Or even worse, “We are prohibited from using AI.”
The optimistic scenario is that AI distributes expertise: a new employee learns faster; a small nonprofit conducts analysis it could not afford; a local business competes more effectively; a person without elite credentials gains a powerful intellectual assistant. The darker scenario is that access to secure tools, training, agents, and integrated systems concentrates among already advantaged firms and professionals.
Leaders help determine which scenario occurs.
3. They are a leadership and governance issue
A ban is not a strategy. A license is not a strategy. A prompting webinar is not a strategy. An agent pilot is not a strategy.
An AI capability strategy requires leaders to answer hard questions: Which tools are approved? What data may be used? Which tasks should remain human-only? Which employees are receiving training? Who has access to agent-building capability? How are outputs checked? How are actions authorized, logged, and audited? What new measures of quality, equity, capacity, and risk will we use? Where are humans indispensable because judgment, trust, care, moral responsibility, or public legitimacy matter?
The ladder is therefore not a call to put every worker and task at Level 5. The appropriate level depends on the work. It is a call for every senior leader to understand the ladder, make intentional decisions, and ensure that safe opportunity is not limited to a small group of early adopters.
A practical leadership agenda: build bridges across all five divides
Leaders do not need perfect answers before they begin. They do need to stop treating AI adoption as a binary choice. A practical agenda might include seven moves:
- Map where your organization actually sits on the ladder. Survey employees not simply on whether they use AI, but how: prohibited or uncertain; occasional task assistance; trained and evaluated use; custom agents; coordinated workflows. Segment results by role, function, seniority, location, and frontline status.
- Separate protective restrictions from capability exclusion. Define confidential and regulated data boundaries clearly. Offer secure approved tools where feasible. Do not force employees into the false choice between falling behind and violating unclear policies.
- Make AI literacy a universal workplace capability. Provide hands-on training in useful tasks, prompt design, output evaluation, sourcing, privacy, bias, and escalation. BCG’s findings imply that one short introduction is unlikely to be sufficient: sustained instruction and coaching correlate far more strongly with regular use.
- Create role-specific use cases. Senior leaders should not merely announce that AI is available. Managers must help employees see how it improves their real work: fundraising research, customer support, grant writing, sales preparation, program reporting, operations, procurement, financial analysis, stakeholder communications, or policy research. And – of critical importance – maintain and widely publicize an AI Prompt/ Instructions Library to catalogue not only what has worked well, but also those that didn’t work.
- Democratize safe agent-building. Identify employees in different functions who can build and test bounded agents with clear human review. Give smaller units and mission-driven organizations access to methods that are not limited to large technology departments.
- Govern orchestration before it scales. As agents begin to interact with systems and one another, they should require permissions, monitoring, auditable source trails, approval checkpoints, performance testing, red-team reviews (that deliberately probe for failures and unintended consequences before deployment), and clear human accountability. The higher the ladder, the higher the governance requirement.
- Measure opportunity as well as productivity. Ask not merely whether AI reduces costs, but whether employees are learning, whether frontline staff have comparable access, whether smaller teams can deliver higher-impact work, and whether new capability is being broadly shared.
The divide we can still choose to prevent
The most important question is not whether AI will be adopted. It already is. Stanford’s 2025 AI Index reported that 78% of surveyed organizations used AI in 2024, up from 55% a year earlier.[18] The U.S. Census Bureau reported in May 2026 that 17% to 20% of U.S. businesses had used AI in at least one business function during the preceding two weeks, with much higher rates among larger firms and in information and finance.[19] Different measurements produce different percentages, but the direction is unmistakable: AI is diffusing rapidly and unevenly.
The question is whether leaders will allow that unevenness to become a durable capability divide.
The first digital divide taught us that providing devices or internet access was necessary but not sufficient. People also needed affordability, training, support, relevant applications, trust, and the opportunity to convert access into improved lives.
The same is true now – with higher stakes and a faster clock.
Organizations that merely allow AI may trail those that teach people to use it well. Organizations that teach prompting may trail those that design agents to perform real work. Organizations experimenting with one agent may trail those that learn to orchestrate multiple agents responsibly across entire workflows. And organizations that restrict AI without building secure alternatives may unintentionally leave employees and communities unable to participate in the next stage of work.
This is why the new AI digital divide belongs on every senior leader’s agenda. It is simultaneously an economic question, a workforce question, a social-equity question, and a governance question.
The divide is not inevitable. But the ladder is already built.
Selected Bibliography
Anthropic. “How We Built Our Multi-Agent Research System.” June 13, 2025.
Boston Consulting Group. AI at Work: Momentum Builds, but Gaps Remain. June 2025.
Brynjolfsson, Erik, Danielle Li, and Lindsey Raymond. “Generative AI at Work.” The Quarterly Journal of Economics 140, no. 2 (May 2025): 889-942.
Cisco. Cisco 2026 Data and Privacy Benchmark Study. 2026.
Gillespie, Nicole, Steve Lockey, Tim Ward, Alex Macdade, and Greta Hassed. Trust, Attitudes and Use of Artificial Intelligence: A Global Study 2025. University of Melbourne and KPMG International, 2025.
Gallup. Kemp, Andy. “AI Use at Work Rises.” December 15, 2025.
Gallup. Kemp, Andy. “Rising AI Adoption Spurs Workforce Changes.” April 13, 2026.
Gartner. “Gartner Predicts 40% of Enterprise Apps Will Feature Task-Specific AI Agents by 2026, Up from Less Than 5% in 2025.” August 26, 2025; updated September 5, 2025.
Kergroach, Sandrine, and Julien Héritier. Emerging Divides in the Transition to Artificial Intelligence. OECD Regional Development Papers No. 147. OECD Publishing, 2025.
Microsoft. 2025 Work Trend Index Annual Report: The Year the Frontier Firm Is Born. April 23, 2025.
Microsoft. “What’s New in Copilot Studio: May 2025.” June 4, 2025.
Moderna. “Collaboration with OpenAI: Transforming the Way We Work and Innovate Through AI.” April 24, 2024.
Morgan Stanley. “Morgan Stanley Research Announces AskResearchGPT.” October 23, 2024.
OECD. Generative AI and the SME Workforce. November 2025.
OpenAI. “Moderna: Pioneering the Future of Medicine with ChatGPT Enterprise.” 2024.
PwC. The Fearless Future: 2025 Global AI Jobs Barometer. June 3, 2025.
Stanford Institute for Human-Centered Artificial Intelligence. 2025 AI Index Report. 2025.
U.S. Census Bureau. “AI Use at U.S. Businesses.” May 2026.
[1] Nicole Gillespie et al., Trust, Attitudes and Use of Artificial Intelligence: A Global Study 2025 (University of Melbourne and KPMG International, 2025), 70-76; Cisco, Cisco 2026 Data and Privacy Benchmark Study (2026), 3-4, 7. The Cisco survey was conducted in September 2025 among more than 5,200 IT, technology and security professionals with data privacy responsibilities across 12 markets.
[2] Sandrine Kergroach and Julien Héritier, Emerging Divides in the Transition to Artificial Intelligence, OECD Regional Development Papers No. 147, 2025.
[3] Andy Kemp, Rising AI Adoption Spurs Workforce Changes, Gallup, April 13, 2026. Based on a February 4-19, 2026 Gallup survey of 23,717 employed U.S. adults.
[4] Andy Kemp, AI Use at Work Rises, Gallup, December 15, 2025. Common uses were reported in Gallup’s Q2 2025 measurement and described in the December article.
[5] Andy Kemp, Rising AI Adoption Spurs Workforce Changes, Gallup, April 13, 2026. Based on a February 4-19, 2026 Gallup survey of 23,717 employed U.S. adults.
[6] Boston Consulting Group, AI at Work: Momentum Builds, but Gaps Remain, June 2025. Global survey of 10,635 employees; frontline analysis includes 3,537 employees.
[7] PwC, The Fearless Future: 2025 Global AI Jobs Barometer, June 3, 2025. PwC analyzed close to one billion job ads across six continents.
[8] Erik Brynjolfsson, Danielle Li, and Lindsey Raymond, “Generative AI at Work,” The Quarterly Journal of Economics 140, no. 2 (May 2025): 889-942.
[9] OpenAI, Moderna: Pioneering the Future of Medicine with ChatGPT Enterprise, 2024; the case study reports Moderna’s adoption figures and describes the Dose ID pilot.
[10] Moderna, Collaboration with OpenAI: Transforming the Way We Work and Innovate Through AI, April 24, 2024.
[11] Morgan Stanley, Morgan Stanley Research Announces AskResearchGPT, October 23, 2024.
[12] OECD, Generative AI and the SME Workforce, November 2025. Representative 2024 survey of more than 5,000 SMEs in Austria, Canada, Germany, Ireland, Japan, Korea, and the United Kingdom.
[13] Microsoft, 2025 Work Trend Index Annual Report: The Year the Frontier Firm Is Born, April 23, 2025. The report states that it draws on survey data from 31,000 workers across 31 countries, LinkedIn labor-market trends, and Microsoft 365 productivity signals.
[14] Microsoft, 2025 Work Trend Index Annual Report: The Year the Frontier Firm Is Born, April 23, 2025. The report states that it draws on survey data from 31,000 workers across 31 countries, LinkedIn labor-market trends, and Microsoft 365 productivity signals.
[15] Microsoft, What’s New in Copilot Studio: May 2025, June 4, 2025.
[16] Anthropic, How We Built Our Multi-Agent Research System, June 13, 2025. The reported performance comparison is an internal evaluation and should not be generalized to every multi-agent use case.
[17] Gartner, Gartner Predicts 40% of Enterprise Apps Will Feature Task-Specific AI Agents by 2026, Up from Less Than 5% in 2025, August 26, 2025, updated September 5, 2025.
[18] Stanford Institute for Human-Centered Artificial Intelligence, 2025 AI Index Report, 2025. Its organizational use statistic draws on a survey measure and should not be compared directly with the Census measure of recent U.S. business-function use.
[19] U.S. Census Bureau, AI Use at U.S. Businesses, May 2026, reporting Business Trends and Outlook Survey data collected December 14, 2025 through May 3, 2026.
If you haven’t been there, the Beekman is also an incredibly cool hotel, located inside what was originally called Temple Court, the 134 year old building that was one of Manhattan’s original “sky scrapers.” It has been impeccably renovated with a soaring courtyard bar and two very good restaurants. I checked in around 1:30 am, and with several bags and CHUbike, accepted the help of the late night bellman to my room. He was incredibly friendly, very professional and engaged me in a conversation about biking. I learned his name was Odane Small. Born in Jamaica, Odane had recently moved to the US.
That alone would have made up for the frustration I encountered the evening before, but the story gets better. When I arrived at Gild Hall, Manager David Finch greeted me with the news that (a) they had comped my room at the Beekman for Saturday night and (b) they had upgraded me to one of their amazing suites for the remainder of my stay. And when I entered my room I found a note from David along with a huge fruit basket and a very nice bottle of wine. Both totally unnecessary and greatly appreciated.
I have often wondered if there is a common event that gets people to start blogging. I imagine for many it’s a topic or an issue they feel so passionate about that they feel compelled to share their thoughts with others. For a wonderful example of this, see my friend 
Having stumble upon this report, I was surprised that I had not heard of it from traditional news media or blogs, and disappointed that Obama wasn’t using this as a framework for his heath care reform efforts. We used this as one of the primary sources for developing Alan Khazei’s health care policy during his race for the Massachusetts U.S. Senate seat.
On the Khazei campaign, we reached out to our network of friends we were introduced to Dr. Alan Woodward, a former President of the Massachusetts Medical Society and a passionate expert on health care cost reduction. Dr. Woodward turned us onto the successful approaches to medical malpractice reform being successfully implemented by the 
Passionate promoter of all I believe in. Entrepreneur, CEO, Advisor to Loyalty, Political and Social Entrepreneurs, Leaders, and Investors. Father, Brother, Athlete, Activist.
