ey business people working at laptop in office

AI apprenticeships facilitate upskilling in innovative technologies


AI apprenticeships offer a learning approach that can clear the way for responsible AI adoption with the right framework and tools.


In brief

  • AI apprenticeships boost upskilling and secure AI practices.
  • Hands-on learning and mentorship across roles is encouraged.
  • This approach addresses AI risks and fosters continuous innovation.

Artificial intelligence (AI) is rapidly advancing technological innovation, expanding the realm of what’s possible for automation, big data and creative content. For government agencies, the potential use cases to drive operations, make faster decisions and provide a better experience for constituents are endless. With the sudden rise of AI integration, especially with generative AI (GenAI) technologies, there is an urgent need to upskill government workforces for AI to be fully utilized, but also evaluated for its risks. AI apprenticeships offer a learning approach that can clear the way for responsible AI adoption with the right framework and tools.

Adopting medieval approaches for leading-edge technology in AI apprenticeships

In previous eras, where more jobs required hands-on work, apprenticeships would play a major role in the transfer of specialized skills — from carpenters to blacksmiths. Even today, this model is leveraged in industries such as manufacturing, where apprenticeship programs provide on-the-job learning in technical areas to create highly skilled staff and increase productivity. According to one study, 91% of apprentices continue at a job where they receive training, reducing turnover and overall training costs.1 In the apprenticeship format, the teacher or mentor works together with the “apprentice” on a skill or task. Rather than trying to grasp an abstract concept, the apprentice can learn by watching and doing the task directly. The teacher directly models their thought process or approach, empowering the mentee to iterate on a task and build their proficiency.

 

Applying the apprenticeship model to AI learning can help accelerate experimentation and build confidence in the workforce. A mentor or teacher might walk through their approach to using web-based GenAI to generate an outline for a business report. They may demonstrate how they develop the prompt for the GenAI model, explaining their word choices and the data they included. The apprentice can take this example and apply it to their own reports, but also broaden the application for other content or writing needs, expanding possible use cases. This model also moves away from the traditional mode of career development, where knowledge transfers occur from the top down, curated training curriculum or a buildup of skills as an individual progresses. AI expertise may reside across all levels of an organization, often with greater concentration of AI literate individuals in the middle or early part of their career. Incorporating apprenticeships as not just a style of learning but as a leadership mindset within day-to-day work promotes employees to create spaces for modeling skills whenever they collaborate on or assign tasks.

 

Early adopters within your organization will be candidates to become mentors to late adopters, allowing more experienced professionals to educate and train their colleagues. This community-based practice encourages individuals who innovate to lead and encourages a culture of codevelopment and collaboration. Apprentices learn how to apply AI leading practices continuously and within their daily work and in the context of hands-on applications, rather than on the basis theory or policy first. Identifying which groups of adopters your workforce falls into within each of the two groups — apprentices and teachers — will help build the structure of AI apprenticeships that supports the uptake of AI in a secure and responsible manner. The groups will be fluid, as more people become confident and can demonstrate how AI can be applied in tangible use cases and to produce business value. More active users of the technology may sound like a risk, but we’ve found just the opposite to be true if the right governance is in place to enable a culture of innovation at scale.

Security and AI literacy

Fully unlocking the potential of AI apprenticeships, while also ensuring continued security, starts with your workforce. The 2024 Work Trend Index Annual Report reports that while 75% of all knowledge workers use AI at work today, a large majority of AI users bring their own AI tools to work rather than having them provided by their organization.2 While professionals recognize the benefits of using AI, unclear and inconsistent approaches to AI integration create their own potential risks. A foundational learning model with sufficient resources to provide AI tools, combined with the skills to use them correctly, will better position organizations ahead of AI risks. Key to this is leveraging the power of the apprenticeship model and relationship-driven learning. This approach can provide new avenues to securing your AI infrastructure against potential threats, including those associated with GenAI tools, including:

  • Protection of sensitive data: Within (GenAI) tools, particularly those that are open source or web based, there exists a risk of unintentional disclosure of confidential information. Apprenticeship models allow the mentor to instruct the learner about the specifics of using sensitive or confidential government data.

  • Integrity of user interactions: The potential for malicious entities to manipulate the reactive nature of large language models is a concern, as it could lead to the creation of deceptive or damaging content, or the facilitation of exploitative endeavors. Early adopters can develop an understanding of responsible ways to utilize AI tools in their projects.

  • Safeguarding training integrity: The deliberate alteration of data or fine-tuning methodologies to embed vulnerabilities, back doors or biases poses a significant threat to the security, efficacy and ethical operation of AI models. Mentors can help learners understand the potential biases and vulnerabilities that exist in their own unique data sets and use cases for AI, creating a tailored approach instead of one-size-fits-all mandatory trainings.

Addressing these challenges requires more than just theoretical knowledge; it demands practical, hands-on experience and a culture of continuous learning. By integrating apprenticeship models into AI education, agencies can create a proactive learning environment that tackles AI complexities head on while instilling a comprehensive understanding of risk navigation.

Recommendations to establish AI apprenticeships in your organization

Apprenticeships are invaluable to develop learners’ critical thinking skills in the realm of AI and to raise productivity and cost-saving benefits. Organizations that do not invest in upskilling their workforce will stagger in the face of digital transformation and new ways of working. Agencies can take the following actions to move toward an engaged learning culture that leverages apprenticeships:

  1. Prioritize learner needs when selecting AI and GenAI abilities to upskill.

    This can include how AI can be most productively utilized for their role and how likely a learner will implement AI consistently.

    Agencies will need to segment their “learner groups” so that there are varying personas based on roles, responsibilities, career levels and skill sets. This will vary for each workforce, and understanding where they stand will allow agencies to choose and mature use cases where the skills are needed most, and in alignment with the technical capabilities of their employees.

  2. Consider learning and development models that challenge the status quo to rapidly scale emerging technologies.

    Traditional hierarchies that pass on skills and knowledge on a rank basis can prevent innovation with new technologies and the distribution of learning success. Agencies can position their earlier adopters to be visible role models for apprenticeships and encourage the existence of mentors across all levels.
     
  3. Bolster multiple avenues for learners to curate their own learning journey and advocate for mentors at all levels.

    Facilitate a range of learning options so that learners can tailor their development according to their role and background. This can include applied project work, establishing communities of practice around GenAI and dedicated workshops. Embracing the diverse roles of AI adopters and fostering this mentorship can enhance education, aligning it with the evolving demands of AI cybersecurity and safety and ensuring a robust, well-prepared workforce for the future of technology.

The views reflected in this article are the views of the author and do not necessarily reflect the views of Ernst & Young LLP or other members of the global EY organization.


Summary

AI is transforming technology, enhancing automation, data analysis, and creativity. Governments can leverage AI for efficient operations, quick decision-making, and improved public services. The surge in AI, particularly GenAI, necessitates training government employees to maximize AI benefits and assess risks. AI apprenticeships can facilitate ethical AI implementation with proper frameworks and tools.


About this article


Related articles

Driving agency AI literacy utilizing guardrails and frameworks

Recommendations government agencies should consider when driving AI literacy.

How to build a strong foundation for federal AI adoption

Discover the EY federal AI report findings on how to build a strong foundation for federal AI adoption.

How the public sector can prepare for AI in the workforce

AI in the workforce presents a tremendous opportunity for organizations to employ talent in ways they never have before.