|
Resemble AI. Detecting the AI agents among us.
Resemble AI, winner of this year’s Visual 1st Best Technology Award, has launched Resemble Agent Detection, powered by its new RAD-S1 “System One” behavioral model. The tool analyzes how visitors move, type, click, and scroll to distinguish people from conventional bots and disguised AI agents across websites and web apps.
As agents such as Meta’s Muse, Anthropic’s Claude, and ChatGPT navigate real browser sessions with increasingly human-like behavior, traditional network- and fingerprint-based bot defenses struggle to tell them apart from people. In Resemble AI’s own benchmarks against stealth and disguised agents, RAD-S1 detected more than 98% of agent visits, compared with detection rates of 3% to 15% for tools from Cloudflare, F5, and Akamai.
The goal isn’t simply to block agents. By identifying undeclared or disguised agents, sites can apply more granular access controls — directing verified customer agents to appropriate APIs or checkout flows while restricting malicious traffic.
Popsa. Keeps on growing.
Congrats to personalized print products startup Popsa, which says it has surpassed £50 million ($67 million) in annual revenue, serving more than 2.7 million customers across 70 countries.
Initially launched as an iOS photobook app in 2016, Popsa has since expanded to Android and the web, while broadening its range of photo print products.
Dazzle. A personal AI assistant that gets to know you through your photos.
Marissa Mayer — former Yahoo CEO, Google VP and originally Google’s first female engineer — has launched Dazzle, an AI assistant designed to understand your preferences by analyzing your smartphone photo library.
By drawing on your camera roll, Dazzle helps with real-world tasks such as planning parties, organizing outings, and choosing gifts. Mayer hopes its visually intuitive approach will resonate especially with busy women, a demographic she sees as underserved among frequent AI users.
The premise: while conventional AI assistants draw context from emails, calendars, and purchase histories — a camera roll offers a different picture: lived experiences, relationships, personal tastes, and offline memories. Dazzle builds on Mayer’s earlier work in contact management and automated event photo-sharing through Sunshine Contacts and Shine.
But the idea has also struck a nerve. In a recent New York Times interview, Mayer described feeling overwhelmed when her daughter’s teacher asked for three bullet points about what made the 10-year-old special. She turned to Dazzle, which generated a response highlighting empathy, creativity, and a love of animals. Mayer found the summary spot-on.
Some Times readers were less enthusiastic, questioning what it means when AI takes on a task as personal as reflecting on your own child. Which triggers a broader question: when does visual AI-powered assistance become the outsourcing of human reflection?
Persnickety Prints & Persnickety Box. Reunited under same ownership.
Photo industry entrepreneur Chari Pack has officially re-acquired Persnickety Prints, reuniting the full-service photo lab with its sister brand, Persnickety Box, under unified ownership. Chari previously sold Persnickety Prints in June 2023.
The acquisition brings Persnickety’s flagship in-person and online printing lab back alongside the innovative subscription-based mobile photo print app. Moving forward, both entities will operate with aligned leadership while maintaining their signature standard: offering exclusively high-quality silver halide photographic prints.
Photo Imaging CONNECT. Join me there!
Gary Pageau of The Dead Pixels Society is once again hosting Photo Imaging CONNECT (PIC) during WPPI. The conference brings together leaders in printing, services, professional photography, technology, and retail for educational sessions, discussions, and networking.
The event kicks off Sunday, February 28, with an afternoon social at KISS Mini Golf, followed by a full-day workshop on Monday, March 1. Conference sessions and speakers will be announced soon.
Attendance is limited to 75. Reserve your ticket here.
Canva. Teaching GenAI the rules of design.
Canva has launched Canva Craft, its first proprietary image-generation model, developed with researchers from Leonardo.ai, a platform Canva previously acquired.
What’s unique: Craft was trained on millions of human-made design templates from Canva’s content library — including posters, presentations, infographics, menus, and typography-rich assets — rather than predominantly on photographs. The aim is to teach the model not just what things look like, but how effective designs fit together.
Canva highlights three capabilities:
-
Design-aware composition. Training on templates teaches spatial hierarchy, layout relationships, and typography.
-
Two-stage generation. Craft first translates a prompt into a structural blueprint covering object placement, scale, text, and visual hierarchy. It then renders the final visual at resolutions up to 2K, supporting more than 30 aspect ratios.
-
Precision and editability. Canva says this approach improves text legibility and graphic detail while allowing users to edit individual elements without changing the rest of the composition.
TilePix. Expanding its Walmart lineup.
TilePix is expanding beyond standalone magnetic photo tiles, bringing a broader range of nail-free wall décor products to Walmart stores nationwide as well as Walmart.com.
Tracer, the company behind TilePix and featured at Visual 1st, is offering:
-
Magnetic frames in more sizes. New 8 × 10, 11 × 14, and 12 × 12 frames feature built-in kickstands for tabletop display and peel-and-stick magnetic pads for wall mounting.
-
The Super Nail. A hidden, adjustable hanger lets users reposition or straighten frames without reapplying adhesive. It supports frames or lightweight artwork weighing up to 3 pounds.
-
TilePix Mini. This Walmart-exclusive 6 × 6 magnetic photo tile is designed for smaller spaces, including dorm rooms, bedrooms, and apartments.
Customers can add their own prints or order photos through Walmart Photo — making it possible to assemble a custom gallery wall in one trip.
Reuters. Fashion visuals: GenAI’s gains, photo studios’ losses.
As we discussed during the Visuals for Commerce panel at Visual 1st, generative AI is reshaping the economics of fashion imagery. Production-cost savings can exceed 90%, but the shift is also putting pressure on models, photographers, and studios.
A new Reuters report highlights three areas of concern:
-
Expansive contract clauses. Brands including Diesel, Calvin Klein, L’Oréal, and Alix Earle’s Reale Actives have introduced contract language seeking rights to create digital twins, swap faces, alter body proportions, or train AI models on talent imagery without additional compensation.
-
Job losses and studio closures. Zalando closed both of its Berlin photo studios and cut more than 340 jobs. Its subsidiary About You also closed its studio but launched Scayle Studios, an AI-generated model imagery service that is said to reduce production costs by more than 90%.
-
Limited legal protections. Protections remain patchy worldwide. One notable exception is New York State’s Fashion Workers Act, which took effect in June 2025 and requires explicit consent for the creation or alteration of models’ digital replicas. Industry groups, including the British Fashion Model Agents Association, are also developing codes of practice to protect image rights.
But the shift isn’t just about cutting costs. It’s also about matching image production to the pace of online retail. As Julian Jansen, VP of marketing at About You, puts it: “You've always had this super fast world of ecommerce, and this super slow traditional world of photo shoots, it didn't fit together in any way — and now with AI, it's a completely different scale.”
Best,
Hans Hartman
| 
