Episode Summary: Would You Let AI Shop for You?

The Business of Fashion Podcast – October 29, 2025
Guests: Sheena Butler-Young (Host), Brian Baskin (Host), Malik Morris (BoF Correspondent)

Main Theme

This episode explores the rise of AI-powered “shopping agents” promising unprecedented personalization and even hands-off, automated shopping experiences. The hosts and guest Malik Morris discuss what these AI tools can (and can’t) do today, what consumers want from such technology, how fashion industry brands are responding, and what it’ll take for shoppers to trust AI with their style—and their credit cards.

Key Discussion Points & Insights

1. Defining AI Shopping Agents

• What is an AI Shopping Agent?
  • Unlike basic recommendation algorithms, these agents claim to deeply learn an individual’s style, preferences, and occasions by analyzing wardrobe uploads and calendars. They can suggest, purchase, and ship items across retailers.
    • Quote (Malik, 01:41):
      "An AI shopping agent... is supposed to learn the shopper and can act on their behalf. So in theory, these agents can respond to very specific prompts like: ‘buy me a revenge dress for my ex-boyfriend's wedding in Paris...’ and it will find the right option based on what it knows about the shopper's preferences and can even complete the transaction for a person."
• Mainstream tech players (OpenAI, Google) and innovative startups (Vatyr, FIA, Gensmo) are racing to create the most effective agents.

2. How Do Current Platforms Differ?

• General vs. Fashion-Specific Tools (03:56):
  • Big AI search platforms like ChatGPT and Perplexity pull options from the entire internet, responding to granular prompts, but are less personalized.
  • Fashion startups (like Gensmo, Vatyr) require users to upload wardrobe photos and access their calendars, aiming for far deeper style understanding and nuanced recommendations.
  • Some now generate avatars that model outfits for the user.

3. Who Wants AI Shopping Agents? The Target User Problem

• The early adopters are tech-savvy, avid shoppers who enjoy experimenting and are willing to invest effort (uploading wardrobes, etc.).
  • Quote (Malik, 07:31):
    "The early adopters of these AI agents are people who are avid shoppers who love new technology... You have to love the process of shopping to want to do that."
• It’s a classic “build it and they will come” scenario: While only a subset craves this now, startups are betting AI shopping will become standard as self-checkout did in retail stores.

4. Trust, Accuracy, and Consumer Readiness

• AI recommendations today are “maybe 80% there”—that lingering 20% gap undermines trust, especially when money is at stake.
  • Quote (Brian, 09:31):
    "If Fio's recommendations are really great 90% of the time, and the other 10% they're just hilariously off the mark, are people going to look at that and go, hey, this is pretty good? Or... easier for me to just shop for myself?"
• The process to full automation will be gradual: Users will engage in dozens or hundreds of interactions before ever letting an agent complete a purchase on their behalf.
  • Quote (Malik, 11:38):
    "[The AI] is not necessarily ‘buy me a revenge dress right now and I trust you’. It's more like: ‘show me those options’. Then after a while, maybe, ‘can you buy these shoes?’... It's like the 20th or 100th interaction—you have to talk quite a few times to get to that point."

5. Brand Integration and Change

• Some brands now use their own agents (e.g., Ralph Lauren’s “Ask Ralph” on its app) or license AI tools to enhance e-commerce, offering advanced searches and even avatar-based try-ons.
• The ultimate goal: Recreate the personalized expertise of in-store sales staff in the digital realm.
  • Quote (Malik, 16:15):
    "These AI shopping agents are supposed to be about replicating the in-store salesperson... the shopping agents are trying to do [that] in a digitized environment."

6. Will AI Reshape Trends the Way Algorithms Changed Music/Streaming?

• There's speculation that widespread adoption of shopping agents could alter fashion design and trend cycles just as TikTok or Spotify algorithms have flavored entertainment.
  • Quote (Brian, 17:46):
    "The algorithm does steer people towards certain types of content... and it feels like maybe we're on the cusp of something similar happening in fashion."

7. Who Ultimately Wins: Big Tech, Brands, or Consumers?

• While marketed as a tool to empower shoppers, the hosts believe the eventual "winners" will likely be the major platforms (Amazon, Google, OpenAI) that consolidate the best startups and lock in consumer behavior.
  • Quote (Malik, 19:49): "You're probably going to see them getting consumed into an OpenAI or Google or Amazon... Those already huge BMs, they're probably going to be the ultimate winners, to be honest."

8. Accountability and Risk

• When an AI agent recommends a product that doesn't exist or messes up, who is blameworthy—the AI provider or the retailer? Consensus: Much of the perceived risk lands on the tech platforms, not the brands.
  • Quote (Brian, 22:31): "Brands have been so eager to partner with these startups and with OpenAI because all the risk is on the AI companies. I mean we blame their results, we don't blame Walmart for the fake toothbrush."

9. Consumer Behavior and Willingness to Experiment

• Despite technological hiccups, the group agrees that modern consumers are more open to experimentation and tech advances than critics assume; feedback will continue to iterate and improve agent performance.

10. Implications for Influencers

• The group discusses whether AI replacing recommendations will dethrone influencers. Consensus: Both can coexist. Influencers inspire “the dream,” while AI handles the practical transaction and curation.
  • Quote (Malik, 27:37): "The influencer is what's daring me to dream and actually helping me hone my sense of style. The shopping agent is supposed to just regurgitate that to me with the best results when I'm trying to actually buy products."
• Some startups are already partnering with influencers or have influencer-investors to blend the two ecosystems.

Notable Quotes & Memorable Moments

• On AI Shopping Agents and Personalization
  Malik (01:41):
  "It can respond to very specific prompts like: ‘buy me a revenge dress for my ex-boyfriend’s wedding in Paris who cheated on me and his girlfriend is prettier.’ And it will find the right option… and can even complete the transaction."

• On Trust and Adoption
  Brian (09:31):
  "That other 20% keeps you from really trusting it… Are people going to look at it and go, hey, this is pretty good? Or… easier for me to just shop for myself?"

• On Who Will Win
  Malik (19:49):
  "[Big tech firms] are probably going to be the ultimate winners, to be honest."

• On Influencers vs. AI Agents
  Malik (27:37):
  "The influencer is what's daring me to dream… The shopping agent is supposed to just regurgitate that to me…"

• Comic Relief Brian (33:08):
  "And for Malik's ex's wedding that we're all having to go to right now."

Timestamps for Important Segments

• [01:41] – What is an AI Shopping Agent?
• [03:56] – How do fashion-specific AI agents differ from ChatGPT/Perplexity?
• [07:31] – Who are the early adopters? Consumer willingness and comparison to the dot-com boom.
• [09:31] – What level of accuracy does AI need to achieve real adoption?
• [14:09] – Examples of brand-owned agents (e.g., Ralph Lauren’s app)
• [16:15] – AI as the digital salesperson and the business case for brands/retailers
• [18:14] – Who captures the value: Brands, tech giants, or consumers?
• [22:31] – Accountability/Risk: Who takes the blame for AI mistakes?
• [27:33] – Will influencers lose their role if AI makes the fashion recommendations?
• [30:22] – Hosts’ personal thresholds for using an AI shopping agent

Closing Thoughts from Hosts and Guest

• Personal Adoption (30:22–32:42):

  • Sheena: Unlikely to use in its present form; wants tools to “fundamentally feel helpful.”
  • Malik: Openness mostly comes from the ability to discover new brands tailored to his style.
  • Brian: Finds current niche uses (event-dressing, gift recommendations) persuasive, but broad adoption is likely some years off.

• Consensus

  • AI-powered agents could reshape fashion shopping, but major adoption is a ways off. Current platforms are often clunky, and brands/tech companies still need to earn deep consumer trust through cycles of feedback and iteration.
  • The next generation of fashion commerce, driven by persistent AI experimentation and the ever-present influence of trend-setting humans, is already being seeded.

For full context and deeper reading, see Malik Morris’s article: “What It Will Take for Consumers to Let AI Shop for Them” on businessoffashion.com.