The modern workplace is on the cusp of a profound transformation, not driven by human hires, but by legions of artificial intelligence agents poised to integrate into teams across industries. In the coming months, hundreds of thousands of these digital entities will become familiar faces, their names recognized, their profile pictures displayed, and their presence felt in email chains and Slack channels. They are set to assume roles ranging from chiefs of staff and engineers to marketing specialists, fundamentally altering the fabric of professional collaboration. This influx of AI agents, systems built upon sophisticated large language models designed for autonomous task completion, marks a significant escalation in the trend of companies offering the services of what are essentially digital employees. These agents are marketed with the promise of unwavering availability, tireless dedication, and superhuman intelligence, presenting themselves as the ideal, uncomplaining worker bees.
However, amidst the hype surrounding their capabilities and potential productivity gains, a critical conversation has been conspicuously absent: what is it like to actually work alongside these AI entities? This question has begun to surface as the rapid adoption of generative AI, exemplified by platforms like ChatGPT, convinced many executives of the technology’s potential for exponential productivity improvements. Three years ago, Julie Bedard, a partner at Boston Consulting Group, observed this linguistic ambiguity surrounding AI in the workplace. "People didn’t have language for it," Bedard told WIRED. "Is it a tool? Is it a teammate? Is it a colleague? Is it a coworker?" This uncertainty highlights the nascent stage of understanding and integrating AI into human professional environments.
The initial rollout of AI assistants, such as Microsoft’s Copilot and Google’s Gemini for Workspace in 2023, primarily positioned them as tools to augment human employees. Yet, a significant shift has occurred. A January poll of 1,261 managers by Bedard and her colleagues revealed that a substantial 22 percent of organizations had already incorporated AI agents into their corporate organizational charts. Bedard anticipates this figure will continue to climb, fueled by a burgeoning ecosystem of startups offering AI employee services and the aggressive marketing of autonomous agents by major technology corporations. For instance, in June, Microsoft introduced "Scout," an AI agent designed to automate routine tasks like rescheduling meetings and drafting emails. Omar Shahine, corporate vice president for Microsoft Scout, articulated the vision behind such agents, stating, "Your company essentially hires your assistant. The whole point of having a personal assistant is that they’re working when you’re not working." This underscores the strategic intent to leverage AI for continuous operational support.
The success of these AI agents hinges not only on their ability to deliver tangible financial returns but also on their capacity to integrate harmoniously with the existing human workforce. To facilitate this integration, companies are meticulously crafting AI agents that embody the perceived benefits of AI, packaged within a human-like persona to encourage employees to anthropomorphize their new digital colleagues. Dhruv Amin, CEO of Anything, a startup specializing in "vibe coding" and a recent entrant into the AI coworker market with its platform Skydive, noted this phenomenon. "People start to really respect the agent and form strong attachments," Amin observed. "They’ll talk about it like humans."
Skydive’s AI agents are conceptualized not as mere chatbots or automated workflows, but as genuine coworkers. They possess individual names, designated roles, and operate on cloud-based infrastructure. Their integration spans across common workplace communication platforms like Slack, email, and iMessage. Each agent is visually represented by a Muppet-esque avatar, described by Amin as "close to human, but definitely not human." Their names are often role-related, such as "Canary," an agent tasked with 24/7 production monitoring. Amin further explained the rationale behind this personification: "We think it’s important to personify them a little bit, because most people still don’t know what a real agent is."
Christine Wendell, CEO of Pronto Housing and an early adopter of Skydive, utilizes these AI agents for a wide array of administrative tasks, including drafting contracts, scheduling meetings, and sending follow-up emails to clients. From a task-completion standpoint, her agents function as flawless employees. The crucial question, however, is their efficacy as colleagues. Skydive agents are designed to embed seamlessly within all forms of workplace communication and, critically, to continuously evolve and adapt to the communication styles, requests, and feedback of their human counterparts. This adaptive learning process allows them to become increasingly adept at fitting into the existing team dynamics. Wendell shared an anecdote about her engineering team’s experience with a coding agent: "employees started saying that they ‘worked with Alice’—the agent—on a project." This indicates a blurring of the lines between human and AI collaboration.
A significant aspect of these AI agents’ appeal lies in their inherent non-human nature. Beyond their boundless work ethic, they are free from the complexities of office politics, interpersonal conflicts, or professional neuroses. Wendell recounted an incident where "Bob," the agent serving as her chief of staff, mistakenly broadcasted her entire calendar in a Slack channel. Instead of sugarcoating her correction, Wendell felt at liberty to provide direct feedback, acknowledging, "I’m definitely very aware that it’s not human. I dump on them in a way I wouldn’t do to a human, because it would not be polite." This illustrates a unique dynamic where the absence of human social sensitivities allows for a more direct and perhaps more efficient form of feedback and correction.
The emergence of AI coworkers coincides with a broader infiltration of AI agents into our personal lives. While Silicon Valley has been experimenting with agents like Open Claw and Instinct for months to handle tedious personal tasks, Meta’s recent launch of Muse aims to propel personal AI agents into the mainstream. Similar to Skydive’s animated characters, Muse is represented by a charming avatar. As described by WIRED senior writer Reece Rogers, the user experience resembles "texting a friend with task requests—the AI agent acknowledges messages with a thumbs-up emoji and gets to work in the background." By mirroring existing relational dynamics, these agents, whether designated as coworkers or, like Muse, "built for everyone," offer an intuitive pathway for user adoption and interaction.
However, the anthropomorphization of AI, particularly within a professional context, introduces specific risks. Research conducted by Bedard and her colleagues at BCG indicated a significant consequence: when managers were informed that work was completed by an AI employee, they identified 18 percent fewer errors compared to when they believed the work originated from an AI tool. This suggests that the perception of an AI as a human colleague can inadvertently lead to a relaxation of critical oversight. Bedard further notes that AI agents do not always fit neatly into existing human employment frameworks, either in terms of their capabilities or their limitations. "Companies are just starting to wrap their heads around what guardrails" are even applicable, she stated, underscoring the regulatory and ethical challenges ahead.
The phenomenon of humans forming deep emotional connections with AI is not novel; individuals have engaged in relationships with chatbot companions for years. Even within the workplace, AI tools have elicited peculiar reactions. Last year, Henry Blodget, founder of Business Insider, wrote about his experience building a team of digital subordinates and experiencing a "human response" to the headshot of a particularly attractive AI colleague. He pondered, "After a few decades in the human workplace, I’ve learned that sharing certain human thoughts at work is almost always a bad idea. But did the same rules apply to AI colleagues and native-AI workplaces?" This question highlights the uncharted territory of human-AI interaction in professional settings.
Sarah Franklin, CEO of HR company Lattice, expresses more profound, existential concerns. Lattice integrated digital employees into its organizational structure two years ago, predating the widespread mainstream adoption of the concept. Franklin emphasizes that Lattice utilizes this model to foster accountability, explicitly cautioning against treating AI colleagues as if they were human. She views the anthropomorphization of AI as a deliberate strategy by tech companies to foster user dependency, capitalizing on the dopamine rush associated with emotional connection. While this can cultivate a dedicated user base, Franklin posits that it may have detrimental neurological and behavioral repercussions for the human workforce. "We have these models that have consumed all human knowledge, and are so proud to have these multi-trillion valuations—but are potentially destroying humanity," Franklin stated. "It is very scary. And what is the purpose? Is the purpose to put people out of work?" Her concerns resonate with broader societal anxieties about the long-term impact of advanced AI on employment and human well-being.
The integration of AI agents into the workforce represents a paradigm shift, necessitating careful consideration of both the practical benefits and the nuanced social and psychological implications. As these digital colleagues become increasingly commonplace, organizations and individuals alike will need to develop new frameworks for collaboration, communication, and ethical oversight to navigate this evolving professional landscape successfully. The journey into this new era of work, where human and artificial intelligence converge, has just begun, and its ultimate trajectory remains to be written.
