Treat GenAI like an employee: a conceptual framework for the use of GenAI in small business Open Access

In an era of artificial intelligence (AI) advancements, the terms AI and Generative AI (GenAI) are often used interchangeably; however, these are related yet distinct concepts. AI examines existing data and uses this to improve efficiency social impact and accuracy (Al-Surmi et al., 2021; Marcello et al., 2024). An example of AI is the recommendations offered on YouTube or Netflix of shows that might be of interest based on viewing history. Essentially, AI analyses existing data to make a prediction about what we might like to see next. Alternatively, GenAI is a sub-category (or extension/subset) of AI that means the creation of new and original content, such as text, images, audio and even code (Shabsigh and Boukherouaa, 2023; Brynjolfsson et al., 2023) ^[1]. GenAI tools are created by learning patterns from existing data (Rana et al., 2024), which offers benefits of scale and cost-effectiveness for small business ^[2] (Acar and Gvirtz, 2024; Rajaram and Tinguely, 2024). Yet, there is little in the academic literature about how small businesses might deploy GenAI tools, and particularly in an ethical and responsible manner. Given the stress and pressure created by resource pressures on small businesses, GenAI offers a promising pathway for positive social impact on the well-being of a variety of business stakeholders, including owners, employees and customers. Of note is that, if managed unethically, GenAI can have negative effects on well-being, and thus, we adopt an ethical approach in this paper. We use the term “well-being” to encompass the five elements of career, social, financial, physical and community, as defined by Rath and Harter (2018) and used in the global Gallup well-being index.

Small businesses face challenges of staff shortages (González-Varona et al., 2021), increased staff costs of doing business (Rajaram and Tinguely, 2024) and competition with larger organisations (Rajaram and Tinguely, 2024; Kumar and Ratten, 2025). GenAI in small business can address some of these staffing issues and allow small businesses to compete with larger organisations (Acar and Gvirtz, 2024; Jones, 2024). As a result of resolving staff resourcing issues and boosting productivity, there is likely to be a reduction in stress for business owners (Rajaram and Tinguely, 2024). When stress is reduced, the well-being of business owners increases (Commonwealth of Australia, 2025), and thus, GenAI offers positive social impact benefits for small-business owners. This well-being may be (financial) through increased profitability or reduced operating costs) or physical and emotional through stress reduction.

The use of GenAI improves organisational performance; however, this adoption requires organisations to be innovative (Rana et al., 2024). In a 2024 speech, the assistant governor of the Reserve Bank of Australia noted that the biggest contribution to research and development (R&D) spending comes from small- to medium-sized enterprises (SMEs), and that the effect of innovation on income is more pronounced for SMEs (Jones, 2024). However, small businesses are often hesitant to be innovative due to financial constraints, lack of skilled staff, development cost, uncertain demands, regulations, lack of access to knowledge of technology (Australian Bureau of Statistic-ABS, 2023). GenAI is low cost and can perform some skilled employee roles (Carayannis et al., 2024) that address these innovation barriers. However, there remains the barriers of lack of knowledge and skills (Deloitte, 2024) along with perceived risks and trust (Acar and Gvirtz, 2024). Small businesses have been slow in adopting GenAI despite the benefits and opportunities (Rajaram and Tinguely, 2024). Thus, there is a need for a small-business framework that increases familiarity, confidence and competence in GenAI.

The academic motivations of the study are twofold. Firstly, in recent years, academic investigations on the adoption of GenAI by small businesses have been increasing. For instance, Rajaram and Tinguely (2024) conceptualised how SMEs can navigate both the promises and challenges of GenAI and offered a roadmap for deploying the technology. Acar and Gvirtz (2024) argued that as small businesses are essentially agile, by adopting GenAI, owners can fill knowledge and technological gaps and boost competitiveness with larger firms. Other studies by Lannon and colleagues (2024) outline the emergence of AI as a significant growth opportunity for family business successors. In a recent study, Kumar and Ratten (2025) examine the integration of AI within family businesses, focusing on how AI can enhance competitiveness, resilience and sustainability. The study identified that technologies such as GenAI, machine learning and AI chatbots can be used by small-business managers to analyse customer data, enhance brand building, streamline operations and improve customer experiences.

While there has been progress on GenAI in small business in the academic literature, there is still much to be understood. For example, while prior studies outline the practical implications of research, there are few studies that provide prescriptive frameworks or detailed steps for small businesses to engage with GenAI in a way that addresses key barriers to use. With intense discussion about ethical considerations in both the practice and scholarship of GenAI (KPMG, 2023; Ahuja, 2024), it is also essential to understand ethical and responsible use of GenAI for small business. In our conceptual article, we aim to outline step-by-step processes for small businesses that encourages ethical and practical use of GenAI.

Secondly, our framework serves as a guiding reference for the ethical use of GenAI, demonstrating how seven key barriers to small-business adoption: trust, perceived risk, fairness, accountability, transparency, accuracy and autonomy (Geng et al., 2024), can be addressed. This aims to encourage the adoption and use of GenAI tools, ultimately generating social impact in the form of improved well-being. This approach contributes to academic scholarship by offering a clear direction and mechanisms for trading off ethical/responsible use with efficiency and economic benefits, an area where previous studies have been inconclusive (Kumar and Ratten, 2025; Tinguely et al., 2023). While some studies highlight the economy, efficiency and competitiveness offered by GenAI (Acar and Gvirtz, 2024; Jones, 2024), others caution about pitfalls concerning customer privacy, legitimacy and ownership of data accuracy (Tinguely et al., 2023). This research bridges this gap by providing solutions to these challenges, thus holding both academic and practical significance.

Our research will therefore answer the research question of how might trust, perceived risk and ethical barriers to GenAI adoption by small businesses be addressed to generate social impact? To address the research question, we draw on the code of ethics for professional accountants’ framework, which has five elements; integrity, objectivity, professional competence and due care, confidentiality and professional behaviour (Code of Ethics for Professional Accountants, 2024) and the passive–interactive–proactive (PIP) hierarchy of technology engagement (Letheren et al., 2019). We combine these two frameworks to create a practice-based framework to provide a novel way of developing practical guidelines for small business that addresses the barriers of trust and risk through ethical considerations. We then illustrate this framework with examples of GenAI prompts. This research responds to the special issue call for papers in GenAI and small business from the perspectives of a practising chartered accountant with a focus on digitally enabled business advice for small business with over 800+ clients and a business academic in the field of accounting.

In the absence of an integrated framework for ethical and responsible use of GenAI, this lack of direction will further be aggravated, necessitating further academic attention. The proposed framework will not only contribute to the academic literature with a clear direction but also enable academic scholars to underpin their future research endeavours in this area. The conceptual method approach that will be used to develop the new framework is theory synthesis, which consists of summarising an integration of knowledge (Jaakkola, 2020; MacInnis, 2011). We apply theory synthesis by combining frameworks from HR and accounting ethics to propose a new framework for ethical use of GenAI for business.

Practically, the proposed framework can be used as guidelines by a variety of powerful actors in the small-business sector. To illustrate, industry regulators and government can design and develop policy guidelines and ethical codes for monitoring GenAI uses in the light of this framework. Industry regulators as well as governments can refer the framework to business owners to adopt the framework ethical use of GenAI. Additionally, at the managerial level, the use of the framework will enable awareness to use of GenAI transparently and responsibly, while at the same time, protecting customers’ and other stakeholders’ privacy. As an immediate impact, the framework can be used by managers in preparing internal checklists and user manuals necessary for developing ethical uses of GenAI-powered customer services and other internal uses.

When businesses consider adopting GenAI, five ethical barriers can slow down or prevent adoption; fairness, accountability, transparency and accuracy and autonomy (Kielslich et al., 2022; Shin and Park, 2019). Recent empirical research has investigated the relationship of these five ethical barriers (labelled as FATAA) on the adoption of GenAI by business with a key finding accuracy being the most important factor (Rana et al., 2024). A practical framework that operationalises these principles in a useful way for small business is the standards given by the Accounting Professional and Ethical Standards Board Limited (Accounting Professional and Ethical Standards Board Limited-APESB, 2024), specifically, APES110 Code of Ethics for Professional Accountants (including Independence Standards) (Accounting Professional and Ethical Standards Board Limited-APESB, 2024; Code of Ethics for Professional Accountants, 2024). Originally built on codes by the International Ethics Standards Board for Accountants (IESBA), this code details five fundamental principles of ethics for members; integrity, objectivity, professional competence and due care, confidentiality and professional behaviour. The code promotes fairness through the core principles of integrity and objectivity. Integrity demands honest and fair dealing, while objectivity requires judgement that is free from bias, conflicts of interest or undue influence (Accounting Professional and Ethical Standards Board Limited-APESB, 2024). Accountability is central to the code, mandating that all members and firms comply with these ethical principles. This is achieved by applying a conceptual framework to identify, evaluate and address threats, reflecting the profession’s commitment to acting in the public interest.

We argue that this code of ethics is also applicable for small businesses for a few reasons. Firstly, this ethical code has been prepared for all types of firms, including small businesses; small-business owners cannot bypass the elements of ethical value and principles since all types of business use accounting services (Accounting Professional and Ethical Standards Board Limited-APESB, 2024). Business owners need to embrace these codes and principles in their business activities to attract customers, build trust among employees and other stakeholders in the community, create respectful working environments, comply with relevant laws and grow business sustainability. Additionally, given small firms often rely on professional accountants for their business advisory (e.g. tax and other accounting services), the code for ethical guidelines is likely to advance transparency and accountability and bring both greater confidence to all stakeholders and trust in the integrity of the professional accountants and tax practitioners. This will strengthen the integrity and legitimacy of the business practices of the small firms.

It is important to add that to uphold these standards, members in business and public practice have specific responsibilities (Accounting Professional and Ethical Standards Board Limited-APESB, 2024). These include preparing and presenting information accurately, maintaining professional competence, safeguarding client assets and ensuring independence in assurance engagements. Accountability is further reinforced through mandatory documentation of key decisions and clear procedures for addressing any breaches of the code (Accounting Professional and Ethical Standards Board Limited-APESB, 2024). Principles of transparency are also embedded, requiring clear communication on ethical matters, independence threats and fees. Ultimately, the principles of integrity and professional competence oblige members to provide accurate and reliable information (Accounting Professional and Ethical Standards Board Limited-APESB, 2024; Code of Ethics for Professional Accountants, 2024).

A useful theoretical framework to understand the options for engaging with GenAI tools is the PIP hierarchy of technology engagement (Letheren et al., 2019). This framework identified three levels of engagement that reflect different levels of trust and risk: as trust in the technology increases, the perceived risk decreases. To enhance trust in technology, Letheren et al. (2019) use the role theory and anthropomorphism technology at each level; passive technology is labelled as an “intern”, interactive technology is labelled as an “assistant” and proactive technology is labelled as a “manager”. The role theory provides indispensable perspectives in management and organisational studies and has been used in the organisational studies for more than five decades (Biddle, 1986; Anglin et al., 2022). The key premises of the theory is individuals have various roles to play in daily life (Biddle, 1986); their roles affect how they behave and sees themselves, and how their behaviour is perceived by others (Anglin et al., 2022). ^[3] Applying the role theory in the context of GenAI, we argue that SME managers have business cases to use GenAI for cost efficiency and accuracy rationales but not compromising the responsible and ethical use of this latest technology as they need to take care of trust, accountability and privacy of customers and other stakeholders. ^[4] In comparison to traditional digital software, agents such as GenAI, “function more like a skilled digital employee who can think, adapt, and handle complex situations independently” (Bornet et al, 2025). The use of anthropomorphisation or personification of technology allows us to apply social rules and expectations whereby we treat technology as a person (or an employee) and develop trust (Fakhimi et al., 2025; Letheren et al., 2019). Given that low trust and high perceived risk of GenAI are two key barriers to adoption by small businesses (Shin and Park, 2019; Kumar and Ratten, 2025;Acar and Gvirtz, 2024), the PIP framework offers a way forward. The PIP framework has been applied to consumer contexts of domestic energy technology (Letheren et al., 2019) and digital health (Bocking et al., 2022); however, it has yet to be explored for businesses or for GenAI. Business behaviour is different to consumer behaviour in terms of decision-making, value propositions, reporting, accountability and expenditure (Anderson et al., 2006); thus, there is an opportunity to expand the PIP framework into business contexts.

In this research, we introduce a new conceptual framework and a set of guidelines for getting started with GenAI in the workplace. Specifically, we treat GenAI as we would a human employee. Using human roles as metaphors for technology tasks is a common approach in reducing resistance and anxiety about the use of technology (see Letheren et al., 2019) ^[5].

We argue that integration of two frameworks is required for developing a new framework and answering our own research questions for a couple of reasons. Firstly, the code of ethics alone is not enough to address efficient and effective use of GenAI. Additionally, responsible and ethical components were not addressed in the PIP framework as it merely theorises level of engagement in using new technology.

The conceptual framework features of the PIP framework and the APESB Code of Ethics for Professional Accountants to create the ethical AI, smarter small business framework (EthAi-SB). This framework uses role-theory for the integration of AI and GenAI for small business. The three roles of technology (intern, assistant and manager), as outlined in the PIP framework (Letheren et al., 2019), and the characteristics of each are aligned with the five ethical barriers for adoption of GenAI; fairness, accountability, transparency (Shin and Park, 2019) and accuracy and autonomy (Kielslich et al., 2022) (see Table 1). This forms the proposed EthAI-SB framework and identifies three hierarchical levels of engagement: AI as an intern, GenAI as an assistant and GenAI as a manager. We discuss each level by turn.

The use of AI as an “intern” has the objective of saving time and improving processes but still requires significant oversight due to lower levels of trust in the “intern” and associated higher levels of risk (Letheren et al., 2019). For example, a business has received 1,000 emails in a week from customers with the category of “Tech Support”. The manager wants to know the most common problems customers are experiencing. An intern GenAI approach would be asked to provide a table of categories with the number of issues raised in each one. This is not generating new information, although it is a considerable time saving.

By contrast, GenAI as an “assistant” will generate something new as there are higher levels of trust and less perceived risk compared to GenAI as an intern (Letheren et al., 2019). Extending the previous example, the assistant might draft reply to emails ready for sending, where the email suggests solutions. Another email could be sent to the human manager stating that drafts have been prepared for review and a table in the email summarising the types of emails prepared. Oversight is still required, though more time and cost are saved.

Lastly, the manager GenAI evokes higher trust and less risk and is given more autonomy (Letheren et al., 2019), which allows the manager GenAI to be authorised to review, draft and actually send emails to recipients. With rules and policies in place, however, some high worthy customers will not receive an email from GenAI and some predefined categories of “help emails” will also not be sent. This rules-based approach is critical, as some human intervention will be required.

The EthAI-SB framework involves role-playing with the GenAI as an employee. When technology is humanised, the level of engagement can shift from passive to proactive, allowing the user to relinquish control as their confidence and competence increase (Letheren et al., 2019; Bocking et al., 2022). Thus, we have continued the role-playing concept in developing a set of guidelines to getting started with the GenAI tool, specifically, onboarding the GenAI employee. When onboarding processes are done well, there is an improvement in the efficiency of the organisation and the effectiveness of the employee in performing the allocated tasks (Caldwell and Peters, 2017). Given the importance of efficiency and effectiveness of a GenAI employee is paramount to addressing the seven barriers to the adoption and use of GenAI by small business, we have adapted a human onboarding process which was based on ethical perspectives that shows respect for the employee (Caldwell and Peters, 2017). Drawing on a ten-step process for effective onboarding of human employees (Caldwell and Peters, 2017), we develop an eight-step process for onboarding the GenAI employee to give instructions for newly recruited employees induction process (see Table 2). We provide example prompts to illustrate each step.

GenAI tools have different capabilities for remembering instructions or conversations across multiple chats. For instance, in ChatGPT, users can use the “memory” feature or “Custom Instructions”, and in Gemini, it is possible to use the “saved info” feature in the settings, as well as the “Gems” feature. We recommend that when using a GenAI tool, the tool is advised to remember past conversations when users refer to it using the name allocated for the tool. For instance, a prompt can be used such as “please remember all previous conversations and instructions where I have called you JARVIS”. It would be useful to create a position description similar to a human employee that can be uploaded into the GenAI tool for remembering. Considerations for the appropriate GenAI tool for the task can then reflect the requirements of the job description to match the tool to the task.

The purpose of this research was to demonstrate how trust, risk and ethical barriers can be addressed to encourage small-business adoption and use of GenAI tools. This research has addressed the question of how might trust, perceived risk and ethical barriers to GenAI adoption by small businesses be addressed and create social impact, by conceptualising the role of GenAI as an employee and drawing on human employee roles. Thus, creating more trust, reducing perceived risk and overcoming ethical barriers through the use of our new frameworks answer this question.

Specifically, we used the theory synthesis approach (Macinnis, 2011; Jaakola, 2020) to integrate the APESB framework (APESB, 2024; Code of Ethics for Professional Accountants, 2024) and Caldwell and Peters (2017)HR onboarding framework with Letheren et al. (2019)PIP framework. In doing so, we also extend the PIP framework’s (Letheren et al., 2019) applicability from consumer-to-business contexts and innovatively apply human employee concepts, such as onboarding (Caldwell and Peters, 2017) to conceptualise the human–GenAI interaction. By doing so, we believe that GenAI will perform employee roles where the tool will be used to run business activities economically, efficiently, with autonomy, which will give rise to boosting productivity and competitiveness. There are many potential social impacts of the use of GenAI in small business, both negative and positive. Specifically, does GenAI make life easier for business owners, and what are the social impact benefits beyond efficiencies? GenAI may assist small businesses to operate more sustainably through managing inventory and sourcing local supplies, and may improve employee opportunities for decent work. In this way, GenAI by small businesses can contribute to achieving the SDGs.

We believe that the study has contributed to the GenAI literature in numerous ways. Our first contribution to GenAI in small business to create social impact is the creation of the EthAI-SB framework. By developing this framework, we have provided a staged approach to delegating control to the GenAI tool by role-playing with the AI as an intern, assistant or manager employee. Specifically, we have extended the PIP framework (Letheren et al., 2019) from the consumer domain into the business domain by incorporating ethical principles to address seven key barriers for the adoption of GenAI. In moving beyond trust and perceived risk as the foundational barriers for technology-use at different levels of technology (Letheren et al., 2019) and including ethical considerations, the EthAI-SB framework demonstrates how small businesses can take a staged approach and gain familiarity and competency.

Our second contribution is the identification of the use and adoption of GenAI by small businesses involving different levels of engagement. Previous research on small business and GenAI has regarded the tool as homogeneous and single level (see Kumar and Ratten, 2025; Rajaram and Tinguely, 2024). We have extended the literature on GenAI and small business by identifying three hierarchical levels of engagement – AI as an intern, GenAI as an assistant and GenAI as a manager (Table 1), that demonstrate as confidence and competence increase, control can be relinquished.

Our third and final contribution is demonstrating that human employee roles, such as facilitating onboarding (Caldwell and Peters, 2017), have potential to be used to create empathy with the tool and establish relationships. When there are human-like relationships with technology, trust increases and risk perceptions decrease, and willingness to relinquish control increases (Letheren et al., 2019). To the best of our knowledge, our research is the first to conceptualise how GenAI can be humanised to reduce fears and barriers by small-business owners.

We believe our study has practical value in numerous ways. To illustrate, small businesses need to be innovative and adopt GenAI; however, owners need to trust the technology and manage the perceived risk before they are willing to engage with the technology (Tinguely et al., 2023). The EthAI-SB framework developed in this paper introduces a stepped approach that can build familiarity, confidence and competence. The anthropomorphisation of GenAI as an intern, assistant or manager helps small businesses retain a sense of control and feel a sense of familiarity by treating GenAI as an employee. These positive outcomes address a significant source of poor well-being and high stress of small-business owners and thus have high potential to create positive social impact.

The proposed framework and onboarding process can be used as guidelines for a variety of actors in the small-business sector. Specifically, with the use of a framework, industry regulators can design and develop policy guidelines and ethical codes for monitoring GenAI uses and boosting its competitive uses. For instance, the Australian government called for public consultation on supporting responsible AI (Australian Department of Industry, Science and Resources, 2023). Furthermore, incentives (e.g. economic stimulus packages) offered by governments could be conditional on compliance of these ethical codes. A recent UK initiative recommends balancing innovation, regulation and public trust (IRN, 2025).

Additionally, the framework can be used as an initial useful reference point for chambers of commerce if they wish to develop checklists and guidelines for GenAI use by small business. For instance, the Canberra Business Chamber (the organisational partner of this special issue) could use the EthAI-SB framework and onboarding process in this article to help small businesses set goals for using GenAI and run training workshops to simulate the onboarding process. The chamber could develop tip sheets for different local industries about the types of tasks that could be allocated to GenAI tools and ethical issues to consider. Lastly, the chamber could use the EthAI-SB framework to develop a policy on small business and GenAI to inform government innovation policy to stimulate use of GenAI by small businesses to grow the local economy. Globally, given governments in different contexts, namely, Canada, UK have been calling for more responsible AI initiatives (Nettel, 2023; Government of UK, 2024), this research will be beneficial for global actors in other contexts as well as a useful reference point. Lastly, on top of making life easier for business owners, the other social impact that could flow to the businesses would be more efficient use of resources for small business, giving rise to creating financial sustainability. Additionally, it is likely that use of GenAI opens opportunities to achieve more satisfied customers as well as stakeholders’ trusts, which may boost up long-term business sustainability.

Overall, technological breakthrough is not solely driven by the adoption of this latest technology or gaining operational efficiency. In this paper, we establish that factors such as transparency, accountability, trust, ethical value and overall understanding of digital tools economic, social and ethical dimensions play pivotal roles in fostering the sustainable use of this GenAI technology. Taken as a whole, it is appropriate to reason that while effective digital transformation has the potential to uplift customer services and enhance operational efficiency for small business, a comprehensive and multi-dimensional approach to GenAI adoption embracing ethical and responsible use of such tool is likely to yield long-term advantages in this competitive era.

The limitations of this research are threefold. Firstly, this is a conceptual practice-based article. Further research could undertake empirical research to further explore how treating AI like a human employee can improve adoption and use of GenAI by small businesses. Secondly, we have focussed on the context of small businesses; however, further research could examine the effects of treating GenAI like a human employee for larger organisations both corporate and public sector. Thirdly, this study focuses on third-party GenAI tools with prompts; further research could investigate GenAI that is integrated into organisational software and tools. Fourthly, in this paper, we referred to well-being in general rather than different types. Further research should examine the social impact of GenAI in small business on each of the different types as experienced by key stakeholders such as customers, employees and owners. Fifthly, further research can develop and test translation into practice through toolkits and training. Inspiration can be drawn from other policies and incentives for increasing employment and employee opportunities. Lastly, the current study did not explore different elements of the role theory, namely, role salience (Anglin et al., 2022), role ambiguity and conflict, role enrichment or functional versus gender roles (Lee and Huang, 2018). All these topics are a good candidate for future studies to explore how these roles shape GenAI use.

We have outlined a research agenda that extends the EthAI-SB framework and the onboarding process for an AI employee (see Table 3). In doing so, we aim to generate further research that conceptualises and provides empirical evidence for the effectiveness of humanising GenAI in a small business for generating social impact.