We are on the cusp of a new wave of hybrid work where organizations won’t just mix in-person and remote workers—they’ll pair humans and AI agents as co-workers. These AI agents will have the ability to take and act on decisions independently and will not be reliant on detailed user inputs, as today’s mainstream GenAI tools are. For example, they will be capable of interpreting context, adapting dynamically to new information, independently ideating, and even partnering with human colleagues to tackle complex and varied tasks.
AI agents are set to go beyond simply augmenting humans to being true co-workers alongside us. By combining human and AI capabilities, these hybrid teams promise to create new possibilities to deliver competitive advantage far beyond incremental productivity gains. This coming shift also demands thoughtful leadership to balance human workers and AI technologies to ensure the unique strengths of each are maximized.
The new hybrid
In large global organizations, many workers already find themselves collaborating through Slack or Microsoft Teams with colleagues they have never spoken to, let alone met in-person. Even with close colleagues, these real-time digital interactions often outnumber face-to-face meetings. Today, there is another human at the other end of those interactions, providing their expertise or performing a specific task. While many workers have already begun incorporating GenAI tools, like ChatGPT, to help with targeted analyses and tasks, the increasing maturity of AI will take this relationship a crucial step further: rather than being a tool or aide to existing human workers, the AI agent will become the “coworker” on the other end of those digital interactions.
This emerging hybrid workforce has been made possible by advances in the natural language processing of large language models (LLMs) that enable humans to communicate with AI agents in the same way they would with a human team member. The reasoning capabilities of LLMs allow natural language instructions to be translated into action without the need for prescriptive code or detailed instructions, or even well-defined steps. Inputs can be more notional, and the AI coworker can still develop and execute a plan, coming back for feedback as needed.
In many ways, the interactions of humans and AI colleagues will be analogous to human passengers in self-driving cars. The cars require a destination, but not specific instructions on when to brake or accelerate. Self-driving cars plot a course, but also receive new data about their surroundings, processing it to plan and execute actions. AI coworkers will be able to act similarly: interpreting context, interacting with other tools and external systems to develop a plan, and even making certain decisions autonomously. They will also maintain task memory so they can learn and improve on the jobs they do regularly.
Moreover, this human-AI co-worker relationship will likely evolve such that AI coworkers can give their human colleagues recommendations, suggesting tasks and even guiding them through them. While this will not happen immediately, along with both technological and human mindset changes, ultimately, both human and AI coworkers may “manage” one another, depending on the context, in ways analogous to current human teaming models.
The ongoing trajectory towards AI agents as coworkers is clear. Startups such as the AI software company Artisan are already developing AI sales agents that go beyond simple automation with hyper-personalized outreach that can even incorporate emotional intelligence in its messaging. The agent continuously learns and adapts, conducts strategic lead research, and proactively manages email outreach. Human sales teams partner with Artisan’s AI agent to define the initial campaign parameters, monitor performance and provide feedback in real time.
AI agents are increasingly being developed for a range of functional roles—including finance, HR and supply chain management—taking on high-level tasks that demand both data-driven analysis and contextual decision-making. The technology’s capabilities will ultimately flip the software-as-a-Service (SaaS) model: AI agents will now be providing service as a software.
Managing the next frontier
This new frontier will be defined by the many novel interactions between humans and AI, resulting in numerous opportunities and technologies. While we cannot know all the possible outcomes of this fast-approaching future, it is nevertheless critical for leaders today to rethink how their teams get work done, and which parts of the workflow are best conducted by AI coworkers or human team members. This will require a creative reimagining of processes to get the most out of the latest technological developments, as well as an understanding of the complexity, risk, and stakeholder impact to balance technical capabilities with business risks.
As NVIDIA co-founder Jensen Huang said recently, “in a lot of ways, the IT department of every company is going to be the HR department of AI agents in the future.” Some companies, like the pharmaceutical company Moderna, have already begun this organizational shift by incorporating digital technologies under the remit of their chief human resources officer or chief people officer.
As leaders prepare to adapt to this coming reality, there are four key areas of change that will be critical to manage:
Building trust: Just as we sometimes see public hesitancy about adopting new AI solutions—like the use of self-driving Waymo cars on the streets of San Francisco—there will surely be early trepidation about having an AI coworker. That’s why trust is a key barrier to widespread adoption. Leaders will need to empower their teams to learn when to trust AI and how to validate or challenge outputs to ensure that final decisions align with overarching business goals.
Early adopters, such as companies already using Salesforce’s AI support application Agentforce, have discovered the necessity of striking this balance. While they reported success in its ability to successfully streamline customer inquiries, they also emphasized the need to develop their managers’ ability to interpret AI-driven findings, and not just view them as infallible. When more complex situations arise, managers must also be able to navigate the interplay between AI coworkers and human colleagues. Building trust may require initially constraining the capabilities of more autonomous or creative agents until their human co-workers have adapted to the new workplace.
Matching and maximizing capabilities: Human and artificial intelligence capabilities are not the same, a duality encapsulated in Moravec’s Paradox: Tasks that are difficult for humans, like complex calculations or analyzing massive datasets, are easy for computers, while tasks that are difficult for computers, such as social skills, come easily to humans.
This balance is in fact a feature of AI, not a bug: it enables diversity and complementarity that was previously not possible. By bringing together human and artificial intelligence, we can increasingly access “augmented collective intelligence” – insights and capabilities that go beyond what either humans or machines could achieve on their own. The speed at which today’s technology is changing means that leaders will have to consistently re-evaluate and optimize what can best be accomplished by humans, by AI agents, and by humans and AI together to extract full value from augmented collective intelligence.
Enabling scalability: The capacity of AI agents can be ramped up and down as needed, and this ability to dynamically scale part of the workforce will require a systems approach to workforce planning. AI agents will be able to operate around the clock and scale on-demand, and hence what is likely to slow operations down are the interfaces between humans and AI. Managers will need to design smooth interfaces between human and AI workers to ensure seamless operations in the new hybrid workforce.
Redefining ‘fit’: Today’s hiring and performance management systems often emphasize overall “cultural fit” to describe an employee’s ability to work effectively within its established teams. Looking ahead, leaders might also need to incorporate “interaction fit,” where the skills they look for in human team members will include the ability to work with AI coworkers.
Despite these changes, we also hypothesize that some fundamental tenets of management will remain critical to fostering a successful human-AI hybrid workforce.
For example, research has shown that diversity increases the quality of problem solving and increases innovation among human teams. We expect this principle to hold not only when considering the added diversity that AI agents will add to all-human teams but also diversity amongst the pool of AI agents, such as different capabilities or training sets.
In a hybrid workforce, certain “AI roles” will likely be highly specialized, like a financial analyst agent for interpreting monthly financials, while others are generalists. Complementarity will continue to be key, and managers will need to find the right mix of specialized and general AI solutions to work alongside human workers.
Fostering collaboration will still be critical. Collaborating with new AI coworkers, however, could require some explainability at the outset to convey how the tech comes to its decisions. Regular AI review sessions, where human team members examine AI outputs, can help. Employee training on basic AI principles will also be critical to enabling humans to confidently evaluate and collaborate with machine-generated recommendations.
No matter how far the technology advances, it is how organizations manage this new hybrid workforce that will ultimately yield durable competitive advantage. Companies incorporating AI agents as coworkers now will have an edge over competitors in productivity, innovation, and cost efficiency. We are at the start of the new hybrid, and those organizations that start early in adapting their ways of working will build the corporate muscle to experiment, adapt, and adopt the coming wave of AI coworkers.
***
Read other Fortune columns by François Candelon.
Francois Candelon is a partner at private equity firm Seven2 and the former global director of the BCG Henderson Institute.
Theodoros Evgeniou is a professor at INSEAD and a cofounder of the trust and safety company Tremau.
Leonid Zhukov is a vice president of data science at BCG. He is the director of the BCG Henderson Institute’s Technology and Business Lab and of BCG’s AI Institute.
Meenal Pore is a principal at BCG and an ambassador at the BCG Henderson institute.
Amartya Das is a project leader at BCG and an ambassador at the BCG Henderson Institute
Some of the companies mentioned in this column are past or present clients of the authors’ employers.
