As generative AI moves deeper into everyday workplace tasks, a new professional question is emerging alongside productivity: Who is accountable for work produced with AI? Learning scientist Rich Carr argues that workers will increasingly need to demonstrate not only that they can use artificial intelligence effectively, but also that they understand, verify and can stand behind the output.
Carr explores that idea in his new 157-page handbook, “Stake Skills: Owning the Work AI Does for You,” published by Brain-centric, LLC and now available in Kindle and paperback editions, with an audiobook planned.
The workplace debate around artificial intelligence has largely focused on productivity.
Can employees write faster? Can recruiters screen candidates more efficiently? Can analysts automate routine research? Can marketing teams produce more content?
As AI becomes embedded in those workflows, another issue is becoming harder for employers to ignore: how do organizations distinguish AI-assisted productivity from genuine professional capability and accountability?
That is the question at the center of Rich Carr’s new book, Stake Skills: Owning the Work AI Does for You.
Rather than treating AI proficiency as simply another technical skill, Carr proposes a framework centered on documenting the relationship between a worker, an AI system and the resulting work.
His argument is that employees may increasingly be evaluated from two directions. Organizations will want workers who can effectively direct AI systems, while also needing confidence that employees retain the underlying knowledge and judgment required to perform and evaluate the work.
From AI Use to AI Accountability
AI can now participate in a growing range of professional workflows, from drafting documents and analyzing information to generating software, summarizing research and supporting business decisions.
That creates a potential accountability gap.
If an employee uses an AI system to produce a report, who verified the facts? Which portions were generated by the system? What did the employee change? Which decisions were made independently?
Carr’s proposed answer is to make that process visible.
Stake Skills introduces the concept of a third category of professional capability alongside hard skills and soft skills. The term “stake” is used in three senses: what a worker risks, what the worker claims and what the worker vouches for.
The concept shifts the conversation away from simply asking whether AI was used and toward asking whether the professional using it can take responsibility for the outcome.
A Ten-Minute Record for AI-Assisted Work
At the heart of the book is a deliberately simple practice: a weekly ten-minute log.
Workers record what they produced, what AI contributed, what they verified or rejected and which decisions remained human.
The approach is designed to create an ongoing record rather than relying on AI detection software after a document has already been produced.
That distinction matters because AI detection has become a contentious area of workplace and education policy. A detection score can attempt to estimate whether content was AI-generated, but it does not necessarily establish whether a professional understood the material, verified it or exercised appropriate judgment.
A process record approaches the problem from a different direction.
Instead of asking “Did AI write this?”, an employer could ask “What did you do with the AI output, and can you stand behind the final result?”
AI Proficiency Is Becoming a Hiring Consideration
The debate comes as employers increasingly consider how AI capability should factor into hiring and workforce development.
The source material cites a KPMG and University of Melbourne survey of 48,000 workers across 47 countries, which found that 57% of respondents conceal their AI use at work. It also cites a Gartner prediction that by 2027, 75% of hiring processes will include testing for AI proficiency.
These developments point toward a changing definition of workplace competence.
AI literacy may eventually become comparable to digital literacy: an expected capability rather than a specialist differentiator. But proficiency may not simply mean knowing how to operate an AI chatbot or generate a prompt.
It may increasingly include knowing when to use AI, when not to use it, how to verify its output and when human judgment needs to override an automated recommendation.
The Human Layer Becomes More Important
That has implications for HR leaders.
Organizations adopting generative AI need policies that go beyond whether employees are permitted to use particular tools. They also need frameworks for determining acceptable use, verification standards, data handling and responsibility for errors.
This is especially relevant in areas where AI-generated mistakes can carry material consequences, including recruitment, finance, healthcare, legal services and employee relations.
The emerging workforce model may therefore place greater value on AI fluency combined with professional judgment.
The employee who can produce a document quickly with AI is useful. The employee who can produce it quickly, identify an incorrect claim, explain why it is wrong and verify the final version may be considerably more valuable.
A Book That Documents Its Own AI Use
Carr has also made the production process part of the experiment.
Stake Skills was itself drafted with AI and subsequently verified by its author. The book documents that process in its sixteenth chapter, while a public production log provides a record of the work behind the book and Carr’s continuing published output.
That approach turns the book into a practical demonstration of its central argument.
Rather than presenting AI-assisted authorship as something that should be hidden, Carr advocates documenting how AI was used and where human responsibility entered the process.
The book includes self-assessment instruments and a foreword by Daria Cupareanu, founder of Amplifiers, along with commentary from Dr. Sam Illingworth, author of Slow AI.
What Stake Skills Could Mean for HR
If the concept gains traction, its implications could extend beyond individual workers.
Companies may eventually need new approaches to performance management, professional development and hiring assessments that distinguish between AI-assisted output and independently demonstrated capability.
That does not necessarily mean returning to a pre-AI workplace.
Instead, employers could assess multiple dimensions: whether a worker can use AI effectively, whether they understand the domain in which AI is being applied and whether they can demonstrate responsibility for the final result.
This could also influence corporate training.
AI literacy programs may increasingly teach verification, source evaluation, workflow documentation and decision-making alongside prompting and tool usage.
The Next Workplace Skill May Be Knowing What to Own
The rapid adoption of AI has made productivity easier to measure in some areas while making authorship and accountability more complicated.
Carr’s Stake Skills offers one response: make the human contribution visible.
The broader workplace challenge is unlikely to be solved by AI detection alone. As organizations integrate AI into everyday operations, they will need clearer ways to establish who made decisions, who verified results and who remains accountable when an AI-assisted process goes wrong.
That could make the most important AI workplace skill not simply knowing how to generate an answer.
It could be knowing what to verify, what to reject and what you are ultimately prepared to own.
Market Landscape
The workforce is moving toward a model in which AI becomes an embedded productivity layer rather than a standalone application. Platforms from Microsoft, Google, OpenAI, Salesforce and Adobe, among others, are bringing generative AI into everyday business workflows.
That makes AI proficiency increasingly relevant to HR, but it also creates new questions around skills assessment. Traditional hiring processes can measure qualifications and experience; AI-enabled work requires organizations to consider whether candidates can effectively collaborate with automated systems while retaining independent judgment.
For HR technology providers, this creates opportunities around AI literacy assessments, skills verification, workflow governance and employee training.
The competitive advantage may ultimately shift from simply deploying AI to building a workforce capable of using it responsibly and transparently.
Top Insights
- AI-assisted work is creating a new accountability challenge, requiring employees to demonstrate what they contributed, verified and ultimately accepted responsibility for.
- Rich Carr’s Stake Skills framework proposes documenting AI-assisted work rather than relying exclusively on tools that attempt to detect AI-generated content.
- AI proficiency may become a mainstream hiring criterion, making the ability to direct, evaluate and verify AI output increasingly important.
- HR leaders may need new assessment models that distinguish AI tool proficiency from underlying domain knowledge and independent professional judgment.
- The book documents its own AI-assisted production, offering a practical example of the transparency and verification process Carr advocates.
Join thousands of HR leaders who rely on HRTechEdge for the latest in workforce technology, AI-driven HR solutions, and strategic insights
