Apple Podcasts  ·  Spotify  ·  Buzzsprout

Paper A

Understanding Human–Multi-Agent Team Formation for Creative Work

Hyunseung Lim, Dasom Choi, S.W. Nam, Bogoan Kim et al. — 2026 — CHI '26 (ACM CHI Conference on Human Factors in Computing Systems)

peer reviewed journal article  ·   ·  read now

Key findings

• People initially tried to let AI agents work together on their own with minimal human input — basically hoping the bots would figure it out. This mostly failed: the agents got stuck in unproductive back-and-forth loops without making real creative decisions.
• After experiencing those failures, participants shifted to a model where the human stayed in charge — setting the creative direction and directly assigning tasks to each agent, rather than letting agents coordinate among themselves.
• When multiple AI agents try to coordinate with each other autonomously, they struggle with value judgments and creative direction — things that require human taste and context. AI agents are better at executing specific tasks than deciding which direction is worth pursuing.
• Letting a human 'orchestrate' the AI agents (think: conductor directing a section of musicians) produced better creative output than letting agents run on autopilot. The human's ongoing involvement was the key ingredient, not an optional add-on.

Marketing implications

• If you're building or using a multi-agent AI workflow for marketing campaigns (e.g., one agent writes copy, another critiques it, another checks brand tone), don't just set them loose and wait for output. Keep yourself in the loop to set direction at each stage — the research suggests autonomous agent-to-agent handoffs often stall or drift.
• When assigning roles to AI agents in a creative project, give each agent a clearly defined, specific job (e.g., 'generate headline variations' vs. 'evaluate against brand guidelines') rather than letting them negotiate roles themselves. Clear boundaries prevent unproductive loops.
• Think of yourself as the director of an AI creative team, not a passive reviewer of its output. The value you add isn't just the initial brief — it's ongoing judgment calls about which direction is worth pursuing.

Paper B

Physical Humanlikeness as A Moderator of The Relationship Between AI Influencer Marketing and Purchase Intention

Rahyono Rahyono, Ayu Nursari, Lestari Wuryanti, Reza Hardian Pratama — 2026 — International Journal of Management Science and Information Technology

peer reviewed journal article  ·   ·  read now

Key findings

• When an AI influencer looks and acts more like a real person — through realistic appearance, empathetic language, and personalized responses — people become more likely to buy what it recommends.
• AI-driven interaction on social media (engagement, personalized communication) increases how interested consumers are in making a purchase.
• The human-like quality of an AI influencer strengthens the connection between AI marketing activity and purchase intent — it's not just about the AI being smart, it's about the AI feeling relatable.
• The researchers argue that making AI feel more human is as important as the underlying technology when it comes to driving sales outcomes.

Marketing implications

• If you're using or testing an AI chatbot, virtual assistant, or AI influencer in a campaign, make it sound warmer and more personal — use empathetic language and reference the customer's specific situation rather than giving generic responses.
• When designing a virtual AI influencer for social media, invest in realistic visual design and give it a distinct personality, not just product knowledge — the human feel matters more than technical polish.
• If you're evaluating whether to use a virtual AI influencer versus a human one, consider that audiences respond better when the AI feels relatable, so a robotic-sounding AI may hurt, not help, your conversion rates.

Paper C

Artificial intelligence across social sciences and humanities: The evolution of marketing analytics in the digital era

Md. Nazrul Islam — 2026 — Journal of Interdisciplinary Research in Artificial Intelligence and Society

peer reviewed journal article  ·   ·  watchlist

Key findings

• Marketing analytics has moved through four stages: first, just counting clicks and traffic; then, predicting what customers will do; then, generating content and conversations automatically; and now, a new stage where companies also need to ask whether their AI systems are fair, trustworthy, and explainable.
• AI in marketing isn't just a technical upgrade — it also changes who gets to define what consumers 'mean' or 'want.' The algorithms that sort people into audience segments are making social judgments, not just math calculations, and this matters for how brands treat people.
• People think about AI in marketing through three competing stories: AI as a helpful assistant, AI as a scary threat that erodes human judgment, and AI as something that will completely reinvent creativity and persuasion. Which story dominates affects whether customers and employees trust AI-driven marketing.
• The biggest gap in current marketing AI research is that technical papers focus on what AI can do, while humanities and social science scholars focus on what AI means — and these two conversations rarely meet. The author argues marketers need both perspectives to make good decisions.

Marketing implications

• When your team buys or builds an AI marketing tool, ask not just 'does it work?' but 'who does it leave out, and can we explain what it's doing?' — the paper argues explainability is becoming a business requirement, not just a nice-to-have.
• If you use AI to write ads, segment audiences, or personalize messages, you're making decisions about how people are categorized — treat that seriously and have a human review the outputs for fairness, not just performance metrics.
• Pay attention to how your customers talk about AI. Some will see it as helpful, others as creepy or threatening. Matching your messaging to those attitudes (e.g., 'AI-assisted' vs. hiding AI involvement) could affect trust and response rates.

Apple Podcasts  ·  Spotify  ·  Buzzsprout