The Evolution of AI in Communications From Writing Assistant to Agentic Workforce Strategy
The landscape of digital communication is undergoing a fundamental transformation, moving away from the novelty of generative text toward a more sophisticated integration of autonomous systems. Sirui Hua, the Head of Audience and Analytics at NowThis, is at the forefront of this shift, arguing that the industry must stop viewing Artificial Intelligence simply as a digital writer and begin treating it as a coordinated team of specialists. As the "infrastructure behind the storytellers" at one of the world’s leading social-first news brands, Hua oversees the intersection of data science, audience engagement, and creative innovation. His approach suggests that the true value of AI lies not in its ability to mimic human prose, but in its capacity to manage complex, multi-step workflows that previously required dozens of man-hours to execute.
From Social Media Beginnings to Infrastructure Leadership
Sirui Hua’s perspective is rooted in a decade of evolution within the media industry. His career began in a PR agency in China, where he navigated the nascent social ecosystems of Weibo and WeChat. This early experience provided a front-row seat to the power of direct brand-to-consumer engagement. Upon joining NowThis ten years ago as an associate producer, Hua immersed himself in the mechanics of storytelling—learning how to construct narrative hooks and emotional payoffs that resonate in a fast-paced digital environment.
This background in production became the foundation for his current role in analytics. By understanding how stories are built, Hua was able to interpret data not merely as abstract numbers on a spreadsheet, but as reflections of narrative beats and audience psychology. He views his role today as building the "infrastructure" that allows creators to succeed. By decoding algorithms and identifying shifting audience interests, he creates a framework where storytellers can focus on creative output, confident that the underlying strategy will ensure their work reaches its intended destination.
The Limitation of AI as a Drafting Tool
A common pitfall for modern communicators is the over-reliance on AI for content drafting. Hua contends that if an organization is only using AI to write first drafts, they are missing the vast majority of the technology’s potential. In this traditional model, the human remains the "operator"—responsible for research, file management, report formatting, and distribution. The AI acts as a glorified typewriter, while the human remains bogged down by administrative friction.
Hua advocates for a shift from "operator" to "manager" or "editor." In this paradigm, the human professional sets the goals, establishes the standards of quality, and makes the final decision on whether a product is ready for public consumption. The AI, meanwhile, handles the "middle steps"—the recursive, data-heavy, and time-consuming tasks that bridge the gap between an idea and a finished product. Hua likens using a frontier AI model solely for drafting to purchasing a high-end smartphone and only using it to make voice calls; the hardware is capable of far more than the user is demanding of it.
The Rise of Agentic Workflows in Public Relations
The next frontier in this evolution is the "agentic workflow." Unlike a standard AI prompt that yields a single response, an agentic workflow involves a multi-agent system where different AI "specialists" own specific parts of a campaign lifecycle. At NowThis, Hua’s team has implemented this through a paid social agent designed to manage campaigns on platforms like Meta and TikTok.
This system operates as a cohesive unit of specialized agents:
- The Setup Agent: Handles the initial boosting and technical configuration of social posts.
- The Reporting Agent: Aggregates performance data into client-facing reports.
- The Optimization Agent: Monitors pacing and delivery, suggesting real-time adjustments.
Crucially, these agents do not operate in a vacuum. They gather "context" by scanning the environments where campaign data actually lives, such as Slack threads, PDF contracts, emails, and internal data warehouses. For instance, the pacing agent can check actual delivery against promised deliverables and draft recommended budget reallocations or ad pauses. However, Hua emphasizes a strict "human-in-the-loop" (HITL) policy: nothing is executed without a human reviewing and approving the agent’s recommendations.
Automating the Mundane: Status Reports and Sentiment Analysis
For many communication professionals, the most significant drain on productivity is "status reporting"—the manual assembly of information from disparate platforms to create a recap for clients or stakeholders. Hua identifies this as a prime candidate for AI automation because it requires very little creative judgment.
Similarly, the analysis of audience feedback has historically been a matter of "vibes" or small-sample sets. With the volume of engagement on high-traffic posts reaching tens of thousands of comments, it is impossible for a human to read every interaction. AI agents, however, can process 20,000 comments in minutes, identifying nuances in what is resonating and with which specific demographics. Hua’s rule for implementation is simple: identify the recurring tasks that team members genuinely dislike, automate them end-to-end, and maintain human oversight at the beginning and end of the process.
Supporting Data and Industry Context
The shift Hua describes is backed by broader industry trends. According to the 2024 Muck Rack State of AI in PR report, while 64% of PR pros are now using generative AI, a significant portion still struggle to move beyond basic content creation. Research from Gartner suggests that by 2026, 80% of advanced creative roles will require proficiency in "AI orchestration"—the exact skill set Hua is championing.
Furthermore, the concept of AI as a "Chief of Staff" is gaining traction among technology leaders. Anthropic co-founder Jack Clark has noted that organizations should stop buying AI as software and start thinking of it as hiring thousands of digital staff members. This reframing is essential because, much like a human chief of staff, an AI agent is only as effective as the context and access it is granted. If the AI is siloed from the company’s internal communications and data, its output will remain generic and of limited strategic value.
The Creativity Paradox: Raising the Floor vs. Setting the Ceiling
While AI can significantly increase efficiency, it poses a unique challenge to creative originality. Hua points to a study discussed by Georgetown neuroscientist Adam Green, which found that while AI can make student writing sound more creative on the surface, it often makes the underlying ideas more homogenous.
This leads to what Hua calls the "floor and ceiling" effect: AI raises the "floor" of quality, ensuring that even mediocre work meets a certain professional standard. However, the human’s "taste" and unique perspective are what set the "ceiling." Without human intervention, the industry risks a "race to the middle," where all content becomes technically proficient but narratively indistinguishable.
Chronology and Event Context: Ragan’s AI Communications Conference
The discussion regarding these agentic frameworks is set to take center stage at Ragan’s upcoming virtual AI Communications Conference on July 22. This event comes at a critical juncture for the industry as firms transition from the "experimental phase" of AI adoption to the "integration phase."
The conference is designed to provide communicators with actionable blueprints for AI implementation. Hua’s session will specifically break down how these multi-step research and drafting flows can be automated, providing a roadmap for agencies and in-house teams to move from manual operators to strategic managers. The event reflects a growing demand for technical literacy within the communications sector, as practitioners seek to stay ahead of the rapid advancements in large language models (LLMs) and autonomous agents.
Broader Implications for the Future of Work
The move toward agentic AI has profound implications for the future of the communications workforce. As manual tasks like data pulling and initial drafting become automated, the value of entry-level roles will shift toward prompt engineering, data interpretation, and ethical oversight.
For leadership, the challenge will be organizational redesign. If a team of specialist AI agents can handle the workload of several junior analysts, firms must rethink their hiring pipelines and career development paths. The focus will likely shift toward "soft skills"—empathy, strategic intuition, and ethical judgment—qualities that current AI models cannot replicate.
Personal Philosophy and Strategic Clarity
Despite his focus on high-tech solutions, Hua finds his strategic clarity in a decidedly low-tech pursuit: long-distance running. Living in Jersey City and running along the Hudson River waterfront, Hua uses the rhythmic exertion of the sport to clear "mental clutter." He notes that his best strategic ideas rarely emerge at a desk but rather during the second or third mile of a run, with the Manhattan skyline as a backdrop.
This balance between technological innovation and human reflection defines Hua’s approach to the future. By offloading the "drudgery" of communication tasks to AI agents, he believes professionals can reclaim the mental space necessary for high-level creative thinking and strategic breakthroughs. As the industry moves toward July’s conference, the message from NowThis is clear: the age of AI as a simple writer is over; the age of the AI-driven specialist team has begun.
Related posts:
