The following article was written by Scott A. Snyder, a senior fellow at Wharton, adjunct professor at Penn Engineering, former chief digital officer at EVERSANA, and author of the book Your AI Life: Foresight, Insight, and Advice to Help You Start Living Your AI-powered Life Today.
Despite over $300 billion of projected AI spending by enterprises in 2025 and almost 90% of knowledge workers using AI at work, only 5% of employees say they are using AI to transform their work. In addition, over half of employees do not trust AI systems and worry about AI taking their jobs. Meanwhile, executives continue searching for positive returns on their AI investments while their workforces continue to wrestle with widespread adoption.
If we look at previous technology waves such as mobile and e-commerce, adoption was less about the technology and much more about changing the behaviors of employees to effectively leverage the technology, as cited in my previous article “Real AI Adoption Means Changing Human Behavior.” But shifting behavior only happens when we help employees elevate both their skill (mastering AI capabilities relative to their role) and will (improving their mindset to experiment with and take ownership of AI solutions). While AI adoption is trending upward, and over 80% of executives believe AI is transformative for their business, only 20-40% of U.S. workers across industries are using AI, according to a recent Gallop poll — highlighting a significant gap in expectations versus adoption.
While there has been significant focus on corporate messaging around the importance of AI and the deployment of AI tools and basic training for employees, adoption seems to be stuck in first gear. In fact, most employees are still searching for their WIFM (what’s in it for me?). The motivation still seems more carrot than stick. If an employee shares that they can save a certain number of hours developing software, creating content, or performing market research, will they get any credit? Are they likely to be asked to do more with less, or worse yet, potentially risk losing their job? There have already been examples of workers failing to report time savings from AI due to this very reason.
So, how can companies show employees they are serious about sharing the benefits of AI adoption? To date, very few companies have modified their incentives and reward programs at the leader, team, and individual levels to drive the right behaviors that accelerate AI adoption. The leadership conversation needs to move from, “How do we deploy AI?”” to “How do we incentivize the right AI behaviors?”
Why Incentives Matter for AI Adoption
Leadership expert Ken Blanchard said, “The fastest way to change behavior is to reward the behaviors you want to see.” Yet too many times, incentives and rewards are not aligned with the actions we want employees to take, let alone the long-term outcomes we want them to achieve with AI. What if each employee could feel that learning and applying AI in their role would not only multiply their capabilities and performance, but would also benefit them and the company over time?
To date, very few companies have modified their incentives and reward programs at the leader, team, and individual levels to drive the right behaviors that accelerate AI adoption.
Performance metrics in most companies today ignore AI altogether. Without incentives, employees may view AI as extra work or a threat to their job. Incentives would set a clear signal that using AI matters to the company and explicitly ties AI adoption to career outcomes. To be effective, incentives must be aligned across all levels of the company, from leaders to teams to individuals.
- Leader level: Tie senior leaders’ compensation or objectives to explicit AI usage and outcomes to cast the right shadow for direct reports and the organization.
- Team level: Share AI tools, workflow changes, and deliverables to promote collective adoption from one team to another.
- Individual level: Reward individual employees for using AI tools appropriately, and share experiments, wins, and failures to help shift the culture and mindset.
These incentives must be coupled with strong leadership messaging and behaviors, along with guardrails and policies that ensure responsible AI development and use. While incentives alone will not drive sustained AI adoption and transformation, they are key in facilitating change. Early AI movers, like OpenDoor and Shopify, have started to incorporate AI objectives into their performance measurement systems and are experimenting with different incentives to reinforce and reward the right AI-related behaviors across the organization.
Based on some of the early models being tested, a new approach to incentivizing AI adoption could include the following design elements:
Priming the Pump — Incentives for Jumpstarting AI Use
AI Innovation Prizes — Many organizations already have a mechanism for rewarding those who bring a new idea forward via employee innovation programs, innovation tournaments, or hackathons. But too often, these programs lose steam due to lack of a defined process and support resources, limited leadership commitment and follow-through, minimal time and investment available for employee innovators, and lack of compelling rewards. Given the rapid pace of AI advancement, companies have an opportunity to rethink their crowd-sourced innovation models to keep up with evolving AI opportunities across the company. These employee “AI-novation” programs should include:
- Seed funding and time for developing and piloting selected concepts
- Support from the company’s AI Lab or Community of Practice to make sure individuals and teams have access to the right resources and expertise
- Escalating prize amounts for reaching different progress gates (concept selection, MVP, successful pilot, deployment), with broad recognition from executive leadership at each stage.
Once this model is refined, it can be run on a continual basis and even used within different business areas and functions. For example, IBM offers “blue points” and Walmart provides cash prizes for employees who identify and deploy new AI innovations.
Without critical thinking and domain expertise, AI will accelerate shallow work. AI must be viewed as a tool to improve thinking, not to stop thinking.
Team bonuses for AI impact — In addition to new AI use cases coming from employee innovation programs, there will be teams across the business thinking about new ways to apply AI to improve everything from customer experiences to business operations. To encourage these teams to focus on the development of their AI solution on top of their day jobs, an AI impact bonus pool can be created that includes different amounts based on the projected impact. For example, a $50,000 pool is set for a solution that delivers a $1million benefit for the company over 18 months. Assuming the solution successfully delivers on the impact (measurement is critical), the team gets awarded the bonus and can decide how to allocate it based on individual contributions. This could also encourage teams to reimagine their activities with AI and open up bolder opportunities and benefits. Microsoft ties a portion of OKRs and bonuses for teams that successfully deliver efficiency gains and/or customer impact with AI-enabled workflows, and Schneider offers annual awards and recognition for the most impactful AI automation ideas.
Employee-level Gain Sharing — If we expect all employees to lean into regular AI use in their roles, then we also need to overcome the fear of the unknown. Even though recent reports show that AI can drive an average productivity gain of 3x in most workers, companies are seeing nowhere near this likely due to both non-engagement with AI and non-reporting of gains. To tackle this challenge, companies could treat each employee’s time as capital. If an individual comes up with a better way to do part of their job that saves four hours per week (or about 200 hours per year), they get to take a portion of that savings (maybe 50 of the 200 hours) to invest in experimenting on an AI innovation or an AI education course. This not only increases the likelihood of them reporting their AI gains (and the company capturing some of the ROI) and focusing on where AI can drive real value, but it also creates a perpetual cycle of AI growth. Microsoft provides AI compute credits to individuals and teams that successfully pilot new AI solutions, and JP Morgan provides AI innovation tokens to individuals and teams that deploy AI prototypes. The tokens can be converted into grants for future AI innovation projects and education/career development.
These incentive strategies should be deployed immediately to help accelerate the right AI behaviors and outcomes across the business by rewarding early adopters who can inspire others in the organization to move. They can be reinforced by having employees write their future AI-powered job description, highlighting the menial tasks they could offload and the exciting newer tasks they could take on with the help of AI. They should also be coupled with leadership messaging that includes regular recognition of AI early adopters and innovators across all parts of the business.
Rewarding AI Outcomes – Shifting Existing Incentives
The next set is focused on changing existing compensation structures to incorporate AI related objectives.
Employee AI Incentives – Make sure every employee has some portion of their compensation tied to AI adoption and impact. This should include demonstrating the right AI-related behaviors such as AI education, experimentation, and responsible application of approved AI tools and solutions. It should also include demonstration of both hard skills, such as AI prompting and agent building, and soft skills, such as critical thinking, creativity, and collaboration with both human and AI coworkers to get the most out of their roles. Metrics should focus more on value versus volume and include AI readiness relative to the role (maybe including peer evaluations), AI certifications, involvement in new AI applications, and the resulting business outcomes. A few leading examples: Walmart and Pfizer are using bonus multipliers tied to AI-driven KPIs, and Amazon and JP Morgan are linking promotional opportunities directly to successful AI usage and deployment. Having these incentives as part of the ongoing performance review cycle and bonus payouts will ensure employees place a high value on elevating their own AI capabilities and translating this into impact for the business.
Technology is rarely the barrier to AI adoption — behavior is.
Transparent Leadership AI Incentives — Assign a meaningful portion of leader and manager compensation to AI adoption and impact in their business areas, and communicate this to the broader organization. These metrics could include the number of employees actively using AI, AI readiness of their organization in terms of skills and mindset, number of experiments/pilots leading to deployed solutions, business impact/ROI delivered by AI, and their personal AI readiness as a leader. AI objectives should also encourage a balance of near-term, continual AI-driven improvements with high risk and high payoff AI innovations to fuel longer term business growth. For example, BuzzFeed moved its CEO’s compensation heavily into stock options to align with its pivot to AI, and P&G created a business unit-level AI scorecard linked to bonus pools and future R&D budget allocations. This not only sends a strong signal about the strategic importance of AI to the employee base, board, and shareholders, it also encourages leaders and managers to prioritize their time to include a significant focus on AI transformation.
Compensation for New and Evolved Roles — While every role in the company will change in some way with AI, from line workers to CEOs, a number of roles will need to evolve faster and new ones will need to be created to support your company’s AI transformation. Roles such as AI product leaders, orchestrators, interaction designers, and ethicists may have analogs in the current enterprise, but they will likely be doing very different things and will need to have their compensation benchmarked against these roles in the broader market. In many cases, these will present growth opportunities for your existing workforce. For evolved roles, these will primarily involve subject-matter experts in different parts of your business who will be asked to take on the significant responsibilities of evaluating and overseeing AI solutions as well as being accountable for their output. This may also include supervising teams of agents and being responsible for their work or mentoring less experienced employees on how to challenge/evaluate AI performance over time. It is unlikely these experts will be able to shed enough work with AI to make up for their new responsibilities, so adding compensation will probably be needed to retain this critical segment of your workforce.
These evolved AI incentives should be carefully phased in over time, starting with softer metrics related to AI behaviors near term and then phasing in harder performance metrics like cost savings or revenue gains. In order to reduce the fear and anxiety related to AI, companies should consider giving employees and leaders a grace period of six months to one year to adopt and experiment with AI before introducing hard metrics into their performance reviews. This allows individuals to increase their comfort level with AI before signing up to specific performance objectives for themselves or their teams.
The figure below shows how the different incentives work together in achieving sustained AI transformation in a given enterprise.
AI Incentive Model Design
What Could Go Wrong?
Most companies are still early in their AI transformation journeys but things are changing rapidly, so having a test-and-learn approach to AI incentives will be essential. But simply paying for good AI behaviors will fall flat unless leaders navigate some of the pitfalls that could stall AI adoption, such as:
- Over indexing on cost-cutting versus using AI to create a multiplier for innovation and growth. This sends the wrong message to employees and also undermines trust, which is essential for getting employees to engage with AI more deeply. Several leaders of companies that were too aggressive on AI-driven workforce reductions have later admitted they went too far and have rehired a portion of the terminated employees. To avoid this, AT&T proactively reskilled 100,000 employees in roles most exposed to technology disruption with modern skills (cloud, cyber, AI/analytics), versus having to hire outside of the company.
- Incentives that are misaligned with value — Rewarding tool usage and vanity metrics alone (minutes of use, clicks, prompts, API calls, lines of code) can lead to superficial adoption and short-term focus. Making sure incentives focus on both usage and longer-term impact helps avoid AI noise and “workslop” that dilutes focus on the most impactful AI opportunities. Also, failure to include non-monetary rewards and recognition may overemphasize the financial benefits for employees versus the opportunity to improve their own future career path.
- Creating AI haves and have-nots — If rewards focus only on a small group of AI champions, other employees may disengage. Incentives need to be accessible to the broader workforce and also reward employees that help inspire and mentor others on AI capabilities. This will be critical to build trust across the broader workforce that AI can be a multiplier for every role and enable every employee to do “more with more.”
- Lack of governance and risk oversight — Heavy incentives without guardrails may encourage misuse or ethical lapses. Ensure responsible AI behaviors are embedded in employee objectives and incentive design. This should include the responsible use and oversight of AI agents where errors can cascade without the right monitoring and supervision by employees.
- Rewarding hard skills over soft skills for AI — Without critical thinking and domain expertise, AI will accelerate shallow work. AI must be viewed as a tool to improve thinking, not to stop thinking. Employees need to know how to effectively task AI, how to evaluate AI’s output for errors, when to call in an expert for help, and how to formulate the next prompt for maximum benefit.
Implementation Checklist for Future AI Incentive Programs
- Review the company’s AI ambition and roadmap. Define measurable KPIs for AI adoption, both usage and outcomes.
- Develop new incentives and rewards to accelerate AI adoption, including innovation prizes, team bonuses for new solutions, and employee-level gain sharing for AI impact. Consider peer recognition, leaderboards, and micro-incentives to complement more structured bonus awards.
- Revise existing incentives and rewards at the employee and leader level to drive sustained AI transformation, and create compensation and incentive structures for new and evolved AI roles. These programs should be phased in to allow for experimentation and learning to improve AI readiness ahead of assigning hard KPIs.
- Ensure a strong responsible AI governance model is in place including guardrails, ethics, and monitoring so that incentives don’t lead to risky AI use and deployment.
- Provide on-going communications. Incentive programs only work if employees know them, understand them, and see leadership doing them.
- Monitor and iterate: Use metrics, surveys, and adoption data to tweak incentive designs over time.
Conclusion
Technology is rarely the barrier to AI adoption — behavior is. By using incentives and rewards at the leader, team and individual levels, companies can encourage positive behaviors, embed AI into ways of working, and move from experimentation to impact. But without thoughtful design and careful integration with other key elements of your AI transformation efforts, financial incentives alone will lose their impact and fail to grow and sustain the skill set, mindset, and trust needed to win in the AI future.
