Apple Podcasts  ·  Spotify  ·  Buzzsprout

Paper A

An Analytical Study on the Role of Satisfaction in Mediating Customer Loyalty and AI-Powered Digital Banking

Mohd Arif Hussain, Md. Rahman, M. A. Hussain, Kasturi Bangla Vanitha et al. — 2026 — International Review of Management and Marketing

peer reviewed journal article  ·  open access  ·  read now

Key findings

• Customer satisfaction acts as the key link between whether someone likes AI-powered banking features (finds them useful, easy to use, trustworthy, etc.) and whether they stay loyal to that bank. In other words, AI features alone don't create loyalty — they do so only by first making customers feel satisfied.
• Gender made no difference: men and women responded the same way to AI banking features in terms of satisfaction and loyalty.
• Age did matter: older and younger customers differed in how AI acceptance shaped their satisfaction and loyalty, suggesting banks should tailor AI experiences differently for different age groups.
• Security concerns, technical problems, and limited network coverage were identified as persistent pain points that reduce satisfaction with AI-powered banking, even when other features are well-liked.

Marketing implications

• If you market a bank's AI features (chatbots, smart alerts, personalized recommendations), don't just promote the features themselves — focus your messaging on how they make banking feel easier and more trustworthy, because that emotional satisfaction is what actually builds loyalty.
• Run different campaigns for younger vs. older customers: this study found age changes how people respond to AI banking features, so a one-size-fits-all ad for your app's AI tools will underperform with at least one age group.
• Fix the basics first: security, uptime, and network reliability are the biggest satisfaction killers. Before investing more in AI features, make sure your AI-powered service doesn't fail — a bad AI experience hurts loyalty more than no AI at all.

Paper B

Empowering Vocational Students through AI-Based Smart Digital Marketing Training: A Case Study at SMKN 1 Grati, Pasuruan

Iqbal Ramadhani Mukhlis, Samas Adimisa Mishbah Habibie, Nambi Sembilu, Najwa Fadhilah Nur et al. — 2026 — Nusantara Science and Technology Proceedings

peer reviewed journal article  ·   ·  test this week

Key findings

• After completing a hands-on AI digital marketing training program, all 43 students scored 100% on the post-test — a major improvement from the uneven, lower scores they had before the training.
• Students learned to use tools like ChatGPT and Canva AI to create marketing content, analyze markets, and design promotional campaigns — skills that previously were not part of their school training.
• Both students and teachers said the program was relevant and effective, and that it filled a real skills gap in vocational education.
• The researchers argue this training model could be copied and used at other vocational schools in Indonesia and elsewhere.

Marketing implications

• If you train junior staff or interns in digital marketing, consider a structured hands-on format using ChatGPT and Canva AI for content and campaign exercises — even a short program can move beginners from zero to functional quickly.
• If you run a marketing agency and want to offer training services, this paper shows there is demand in emerging markets (e.g., Indonesia) for AI digital marketing education — a simple workshop format could be productized and sold to schools or small businesses.
• Don't overclaim results from short training interventions; pair any AI skills training with follow-up projects or real campaigns to confirm skills actually stick.

Paper C

Ethical issues of large language models: a multi-level thematic synthesis of the academic literature

Helen Pervez, Matti Minkkinen, Henrietta Jylhä, Matti Mäntymäki — 2026 — AI and Ethics

peer reviewed journal article  ·   ·  watchlist

Key findings

• Three big ethical problems show up in both the AI model itself and in tools like ChatGPT: the AI can be unfair or biased toward certain groups, it can expose private information, and it can make things up that sound true but aren't (called 'hallucinations').
• When people actually use a tool like ChatGPT, extra problems appear — including scammers using it for phishing attacks, and users not being sure whether to trust it or whether it's being honest about being an AI.
• In business specifically, companies may get short-term gains from using LLMs but risk serious problems: employees may over-rely on AI decisions without questioning them, confidential company data may be exposed, and it becomes hard to explain why the AI made a certain choice.
• Ethical problems are like a game of telephone — a flaw baked into the underlying AI model can get worse or change shape as it travels through ChatGPT and then into specific business or healthcare tools. By the time you notice the problem in the final product, it's very hard to fix it at the source.

Marketing implications

• If your team uses AI tools like ChatGPT to write ads, emails, or content, assign someone to spot-check outputs for biased language (e.g., stereotypes about age, gender, or ethnicity) before anything goes live — the AI can unknowingly reproduce these from its training data.
• Don't paste customer data, campaign strategy documents, or client briefs into a public AI tool — this paper confirms that privacy leakage is a real risk at the tool level, not just a theoretical one.
• If your team is using AI to make decisions (e.g., audience targeting, budget allocation), build in a regular human review step so no one blindly follows what the AI recommends — over-reliance is flagged as a specific business-level risk.

Apple Podcasts  ·  Spotify  ·  Buzzsprout