2025 is coming to an end which means it’s almost time for the release of Spotify Wrapped. Spotify Wrapped is a way for Spotify users to get an interactive and nostalgic end of the year recap of their most listened to artists, genre, and songs. They even curate playlists that are extremely personalized to their listeners.

Last year, Spotify Wrapped received some backlash due to the AI generated responses used in their campaign. People were expecting something unique, but instead received something generalized.

I asked our PR & growth experts: Can platforms and brands utilize AI without losing the personal connection with their brand image and users?

Here are the experts' insights.

• Users will sniff out auto-generated and generic work
• Train models on slang, fandom jokes, and chaos
• AI can create the space to engage in more meaningfully ways
• AI should assist teams in better understanding their audience
• AI should make personalization possible, not deliver it.
• Don't let AI touch the personal storytelling
• Automate the analysis, humanize the application
• If the interaction creates the memory, keep it human.
• AI should never touch the why, only the what.
• Put humans front and center of audience-building creative activities
• AI should do the listening, not the talking.
• Utilize AI as a precision tool of data aggregation and pattern recognition
• Involve AI in studying emotional aspects, not just analyzing behavior.
• Personalization is more than inserting a name or finding data
• Connection disappears when efficiency replaces feeling
• Connection weakens when communication feels engineered
• Humans need to do the curating
• The human brand voice has to deliver unique insights
• Let AI find interesting conversation trends

Users will sniff out auto-generated and generic work

Emily Reynolds-Bergh, Owner at R Public Relations

"I’m a staunch believer that AI should be used as a tool, but not a full employee. Especially in written communications for brands and individuals, human copy makes a big difference. Brands should expect their content to be shared widely on the internet, which means users will sniff out auto-generated and generic work. Take the time to do two things: train your AI tools to speak differently to different audiences, and spend ample time editing and refining whatever it spits out. With this method, you can save time and save face."

Train models on slang, fandom jokes, and chaos

Ali Zubairy, Head of EU/UKI at Visionet

"AI gets accused of being soulless, but really, it’s only as bland as the data it’s fed. If you train an algorithm on buzzwords, you get buzzwords back. Spotify Wrapped lost the plot when it used AI to summarise emotion instead of amplify it. The smartest brands in 2026 will train their models on slang, fandom jokes, and chaos that are the human fingerprints that algorithms usually iron out."

AI can create the space to engage in more meaningfully ways

Pilar Lewis, Public Relations Associate at Marketri

"AI is not inherently bad. The backlash comes from when users don’t treat it as an extension of their own insight and instead use it as a replacement for their thoughts. When brands use AI to better understand audience behavior, personalize experiences, or automate repetitive tasks, it creates the space for teams to use that time in engaging in more meaningfully ways with their target audience. The risk, and where we see the negative side of AI, comes when it’s not monitored. That’s when messaging becomes generic, tone is inconsistent, and there’s no meaningful engagement."

AI should assist teams in better understanding their audience

Suhail Patel, Director at Dustro

"I've collaborated with companies that employ AI technologies to customize advertising without sacrificing human interaction. Last year, tone rather than technology was the cause of Spotify's backlash. Instead of seeming like a buddy, Wrapped began to sound like software speaking to you.

The message should never belong to AI. In order for the human voice to be heard more clearly, it should assist teams in better understanding their audience. When the voice praising them for listening sounds fake, people are annoyed. They don't mind when algorithms curate their playlists or highlight their favorites.

The solution is straightforward: let humans handle the meaning and let AI tackle the math. Connection always prevails over authenticity."

AI should make personalization possible, not deliver it.

Jeffery Loquist, Senior Director of Optimization at SiteTuners

"I've spent 18+ years optimizing how brands connect with users, and the Spotify Wrapped backlash nails exactly where companies go wrong with AI. They automated the one moment that needed to feel uniquely human—the emotional payoff.

Here's what we've proven with actual testing: AI should make personalization possible, not deliver it. At SiteTuners, we increased one client's revenue by 14% using geo-personalization—showing a Texas flag to Texas visitors. Simple data, human insight about what resonates emotionally. The AI identified location; we decided what that location meant to people psychologically.

The line is crystal clear in our work—use AI to surface the insight (browsing patterns, abandoned carts, time on page), but let human strategy decide how to respond emotionally. When we helped Irish Jewelry Craft triple their revenue, AI tracked who showed interest in heritage pieces. We crafted the messaging about emotional connection to Irish roots. One's a data point, the other's a relationship.

Spotify had all the data for great personalization but templated the delivery. That's backwards. If your brand moment is the product itself—like Wrapped—AI should be invisible in the background, not the voice your customer hears."

Don't let AI touch the personal storytelling

Jacqueline Rufflo, President at JR Language Translation Services

"I've spent 20+ years localizing marketing campaigns for global brands, and here's what breaks: AI can't tell you that your clever May 2-4 Canada reference means nothing to German audiences, or that Americans want you to cut to the point while Germans expect data first. Those cultural instincts come from humans who live in those markets.

The smartest move I've seen clients make is using AI for the heavy lifting—translation memory systems that keep terminology consistent across 50 markets, or neural machine translation to draft initial video subtitles—then having native translators transcreate the emotional hooks. We recently worked with a SaaS company that used MT to translate their app UI in 12 languages overnight, then brought in our bilingual team to rewrite every onboarding message and error notification. Users in Mexico and Japan both felt like the app was built specifically for them.

Spotify's problem wasn't using AI—it was letting AI touch the part people were excited about: the personal storytelling. When we localize ad campaigns, machines handle the volume (translating 10,000 app strings), but humans own the punchline (transcreating that one tagline that makes someone laugh or book a demo). The rule is simple: automate what scales, humanize what converts.

Your Wrapped playlist algorithm can be 100% AI. The message explaining why that playlist matters to you specifically? That needs someone who understands why finding your #1 song from a breakup summer hits different than generic "you listened a lot this year" copy."

Automate the analysis, humanize the application

Gunnar Blakeway-Walen, Marketing Manager at FLATS

"I manage $2.9M in marketing spend for 3,500+ apartment units, and here's what I've learned: AI works when it reveals patterns that humans then personalize, not when it generates the final message. When I analyzed resident feedback through Livly, AI flagged recurring oven complaints during move-ins. The data pointed to the problem, but our solution—creating FAQ videos with our actual maintenance staff—was deeply human and cut move-in dissatisfaction by 30%.

The mistake Spotify made was using AI at the emotional payoff moment. We do the opposite in multifamily marketing. I use AI and automation for UTM tracking and campaign performance analysis across our portfolio, which increased qualified leads by 25%. But when prospects tour our properties like The Myles in Vegas, they get real stories about the Arts District legacy and community vibe—things no algorithm can authentically convey.

My rule: automate the analysis, humanize the application. I ran digital campaigns through Digible that used AI for audience targeting and bid optimization, which lifted conversions by 9%. Then our leasing teams took those warmed-up leads and had genuine conversations about lifestyle fit. The tech made us smarter; the people made us trustworthy."

If the interaction creates the memory, keep it human.

James Bernard, Owner at Alcatraz Escape Games

"I run Castle of Chaos haunted attraction and Alcatraz Escape Games, and we've learned something crucial: AI works great for logistics, terrible for emotional connection. We use automation for booking confirmations and capacity management, but when someone asks which escape room fits their teen's birthday or if our horror experience is too intense for their group, that conversation happens with a real person who knows our rooms inside out.

The Spotify situation failed because they automated the payoff moment—the emotional reveal people waited all year for. We do the opposite at our escape rooms. AI handles the boring stuff like sending reminder texts and processing waivers, but when you're stuck on a puzzle at minute 45, our game masters give live, personalized hints based on watching your specific team's dynamics and playing style. That's not something you can template.

Here's what actually works: let AI crunch the data, let humans deliver the insight. After escape room sessions, we track completion times and puzzle difficulty automatically, but our party room hosts personally ask groups what moments made them laugh or what scared them most. Those conversations created our most successful room themes—like when multiple groups mentioned loving our Chloe horror room's specific scares, we knew exactly what elements to amplify. The data told us it was popular; real conversations told us why.

The rule is simple—if the interaction creates the memory, keep it human. If it just enables the memory to happen smoother, automate away."

AI should never touch the why, only the what.

Cheryl Cassaly, Vice President of Marketing at Rehab Essentials, Inc

"I work with universities launching hybrid healthcare programs, and we faced this exact tension when rolling out our University Licensing Select platform. We had 200+ hours of pre-recorded content that could've easily been pushed out with AI-generated course descriptions and standardized learning paths—but here's what we learned: AI should never touch the why, only the what.

We use AI to analyze which course modules students access most frequently and identify knowledge gaps across cohorts. But when faculty reach out to struggling students or customize a flipped classroom session, that's 100% human-designed based on real teaching experience. In our 2024 programs, 58% of enrollments came from student and alumni referrals—that doesn't happen with generic AI messaging.

The breakthrough for us was treating AI like a research assistant, not a spokesperson. It helps us spot patterns in accreditation documentation or flag which hybrid delivery models work best for specific university infrastructures. But when we're coaching faculty on how to facilitate deeper learning or crafting narratives for accrediting bodies, those conversations are entirely shaped by people who've lived in those classrooms.

Spotify's mistake was letting AI generate the emotional payoff instead of just organizing the data that leads to it. Users don't want AI to tell them what their year meant—they want their actual data presented in a way that lets them draw their own meaningful conclusions."

Put humans front and center of audience-building creative activities

Rob Edwards, Senior Content Writer at Savvy Law Firm Marketing

"Audiences long for authenticity and connection in their chosen brands. Thoughtful companies use AI to eliminate communications roadblocks to their target audience rather than eschewing tangible connections for short-term productivity gains. Winners of this new era will use AI in analysis and analytics, and put the humans front and center of audience-building creative activities."

AI should do the listening, not the talking.

Deepak Shukla, Founder & CEO at Pearl Lemon PR

"Spotify Wrapped didn’t fail because of AI, it failed because it forgot the soul behind the data. At Pearl Lemon PR, we use AI every day to analyze audience sentiment, but never to speak for us. AI should do the listening, not the talking. The moment you let it write your thank-you note, your customer feels the switch.

Personalization isn’t about clever phrasing; it’s about emotional recall that human wink that says, we see you. Spotify Wrapped worked best when it felt like your best mate teasing you about your guilty-pleasure playlist. The fix isn’t to ditch AI, it’s to train it to whisper, not perform."

Utilize AI as a precision tool of data aggregation and pattern recognition

Jesse Singh, Founder at Maadho

"By confining the use of AI to data and insights but no further, brands can use AI in a way that does not require it to compromise their personal connection in order to have a more effective marketing strategy. My team noticed that Spotify's backlash was related to the audience's sense of a mass produced output in which they were expecting something much more personally specific than that. The answer, however, is to utilize AI as a precision tool of data aggregation and pattern recognition for the five most unique listening patterns and habits or the one hundredth most obscure artist listened to by each user. This intelligent data intelligence helps the brand easily understand the individual user profile at a granular level not possible when done manually. This human creativity and brand ambassadorial relationship has to be preserved through making human creative resource available for development of the framing and end textual output using this insight derived by AI as a central repository. For instance, a brand might be able to use AI to determine that a user listened to a particular niche genre 350 times more than the average user, but a human writer is needed to write the personalized text of a joke that humorously celebrates that hyper-specific fact. This clear separation will ensure the brand's voice is authentic and the relationship between the brand and customer is not just lost in generic robotic language as the AI assists the narrative but doesn't write it."

Involve AI in studying emotional aspects, not just analyzing behavior.

Joel Lim, Finance Expert at Becoin.Net

"AI can take personalization to new heights, but when it takes over the emotional part of the storytelling process, the brand no longer has a voice.

Spotify didn’t miss the point because it relied on AI technology—its main problem was that it overlooked context and tone as the very essence of human communication.

A computer-generated summary can’t compete with human emotions and memories even if it is very accurate, when people are looking for reflections and nostalgia it just feels empty.

The solution to this isn't to steer clear of AI but rather to involve it in studying emotional aspects, not just analyzing behavior.

This will require the models to be trained with brand tone guidelines, voice characteristics, and even sentiment data from customer feedback.

AI has to be a partner in the creative process but not an enabler through automation.

The smartest brands are those who allow AI to increase the level of empathy, with machines taking care of the structure and humans polishing the essence.

When the tale is told in a way that feels real, the audiences do not care whether it was a human or AI who wrote it they just care about the authenticity of the experience."

Personalization is more than inserting a name or finding data

Faraz Hemani, Owner & CEO at Iron Storage

"Running a business that has grown to 32 stores and acquiring properties across the nation has taught me the importance of an authentic connection with people. I see the Spotify Wrapped situation and immediately relate because even with data and technology, if the human element is missing, the experience falls flat. Personalization and genuine engagement are lessons I carry into every part of my work.

The problems with Spotify Wrapped show the fine line that technology can tread with a personal touch. People look at Wrapped because they feel it’s reflective of their year, their tastes, etc. When AI produced its generalized responses, it stripped away the individuality and the whole thing felt like it was manufactured. Audiences like to see authenticity, rather than efficiency. An obviously automated response can feel cold and impersonal, and worse, in this situation alienate people rather than get them excited for the product. Personalization is more than inserting a name or finding data but needs to contain emotion, context and a bit of unpredictability that people add naturally.

Brands can use AI without losing touch with a personal feel, as long as care is taken with it. AI can handle some of the repetitive hoops and put out vast amounts of information, but the human element should be in charge of how the product is presented. A brand could use AI to gather information about people’s tastes and styles, then put those insights into the form of hints, messages that feel intentional, emotional or talking about what they did. It is a matter of including a fine balance where technology adds to the creativity of human endeavor, rather than taking it away. AI can provide the implements, but the nuances and little quirkiness that make interaction feel memorable still need to have a human hand. If well handled, it can make products feel smarter and more engaging without feeling robotic."

Connection disappears when efficiency replaces feeling

Jason Hennessey, CEO at Hennessey Digital

"AI should enhance character, not dilute presence, within branded experience. The best systems anticipate need yet preserve humility during delivery. Audiences respond to care expressed through tone rather than perfection. Connection disappears when efficiency replaces feeling across interaction. Balance restores value through mindful application and restraint.

Brands can maintain connection by framing AI as assistant guided through empathy. Each interaction should reveal understanding shaped by real observation. Technology must echo sincerity, not impersonate emotion, within outreach design. Collaboration between creative and technical teams guarantees consistency of soul. When compassion directs process, loyalty becomes enduring expression of respect."

Connection weakens when communication feels engineered

Marc Bishop, Director of Business Growth at Wytlabs

"Brands using AI often lose warmth because automation neglects interpretation. Data defines preference yet ignores emotion shaping perception. Connection weakens when communication feels engineered rather than felt. True personalisation begins with curiosity about human need. Authenticity becomes outcome when awareness drives creation.

AI can reinforce individuality by reflecting tone informed through real interaction. Instead of replacing creators, it should amplify their intuition across channels. Machine precision combined with human context forms trust through balance. Every campaign must sound lived, not manufactured, through sensory detail and care. Presence triumphs when brands translate empathy into experience faithfully."

Humans need to do the curating

Chris Kirksey, Founder & CEO at Direction.com

"Spotify Wrapped got a little off which goes to show you just how tricky it can be to mix AI with something people care about deeply like music and how it relates to their own experience. The problem, however, was not technology but how it is applied to create an experience. Folks wanted something that felt like a reflective thing that was handcrafted, like a personal message from a true friend who actually knows them. When the AI filling in the personal touches incorporated the same old generic phrases and language, it came off as hollow. Things like Wrapped carry emotional significance that is tied to music and memory, so anything that interferes with that authenticity will alienate the audience.

Corporations can use AI and still retain personal connection to the user, however, but it is important that they view it as a tool for enhancing human insight rather than replacing it. That is to say, while it can work with tedious data work, point out patterns and perhaps suggest ideas, it is to the humans that they must look to shape the presentation stage for the approach. The best thing is if the humans are the ones doing the curating, from the context to the output to the relevancy of their audience. It is when the inputted data is used as an informant to creativity and not as a dictator that the personal aspect becomes more real. The more AI feels it is assisting a human being than speaking for the corporation the more authentic the user’s connection becomes."

The human brand voice has to deliver unique insights

Raphael Akobundu, Nurse Practitioner at Huddle Men's Health

"Brands can absolutely use artificial intelligence, without losing the personal connection, but the strategy requires that the personalization mechanisms are performed by AI, but not the voice itself. Spotify Wrapped's 2024 problem was based on the use of AI to drive generalized commentary that was impersonal rather than limit AI to the data aggregation and unique playlist curation where it thrives. The machine should be able to go through millions of data points to produce truly unique information for each user, such as "You listened to this certain 1980s Bulgarian folk song exactly 57 times between 2:00 AM and 4:00 AM" because that kind of measurable detail is uniquely personal. The human brand voice then has to deliver that unique insight with authentic, high-quality, pre-approved text and visuals. In my practice, we use AI to run the patient's lab trends across thousands of clients, which allows my team to deliver highly personalized health advice that feels very specific to the patient it is addressed to, because it is based on their unique metrics, and not boilerplate text."

Let AI find interesting conversation trends

Isabella Rossi, CPO at Fruzo

"The anger over Spotify Wrapped happened because the AI went too far. It tried to fake a human connection instead of just making the connection better.

The trick is to use AI as a tool for scaling up. AI is great at sorting data to find a personal detail, like your obsession with 80s synth-pop. But the way you celebrate that detail, the witty comment or shared nostalgia, must come from a human creative team. AI should hand you the tools, but it shouldn't try to make the final work itself.

At Fruzo, we use AI to look at online behavior and find interesting conversation trends. But we would never let an AI write a message to a user celebrating their app anniversary. That has to come from a place of real human understanding."

#PRontheGO