23 Jul 2026
· 9 min read · Forensics · Attribution Can you tell which model made an image? Usually yes — just not the way you would want. Attributing an image to its exact source model is nearly solved in the lab and nearly hopeless in the wild. The fingerprint that matters for QA is the one you already have: the model’s name. Read →23 Jul 2026
· 8 min read · Model watch · Text & anatomy The newest image models are marketed on the two things they still get wrong Nano Banana Pro, Seedream 4, Ideogram 3, FLUX Kontext — every launch leads with legible text and photorealism. The benchmarks say those failures just moved to where you are less likely to look. Read →23 Jul 2026
· 9 min read · Regulation · Disclosure Every platform labels AI content differently. Here is the 2026 map. One AI ad, a dozen platforms, a dozen labels and triggers — and most of the automatic ones ride on metadata that does not survive upload. On 2 August, the EU stops leaving it to the platforms. Read →15 Jul 2026
· 9 min read · Case study · IP & provenance The Filippa K case: an AI campaign copied real photographers — and one image search caught it Three weeks after launch, a photographer put the AI campaign next to the originals. Two days later the whole campaign was gone. The check that would have prevented it takes seconds. Read →15 Jul 2026
· 8 min read · Technical · Text rendering Why AI still can’t spell: the technical reason text in images goes wrong Garbled lettering is not a maturity problem that the next model version will outgrow — it is a consequence of how image models are built, and it has a known workaround. Read →15 Jul 2026
· 8 min read · Detection · Provenance “Is it AI?” is the wrong question — what independent tests say about AI-image detectors Detectors score 98% in the lab, lose half their discriminative power in the wild, and accuse real photos of being fake. The useful question was never whether an image is AI. Read →13 Jul 2026
· 8 min read · Regulation · Nordics Denmark wants to give you copyright to your own face. Here’s where the Nordic rules actually stand. The deepfake bill made global headlines. It still isn’t law — but a Nordic marketer publishing AI imagery is already regulated three times over. Read →13 Jul 2026
· 8 min read · Agencies · Contracts Agencies industrialized AI creative. The paperwork hasn’t caught up. Every major holding company now runs an AI production platform. The questions that decide who gets sued — ownership, disclosure, verification — are still handled ad hoc. Read →10 Jul 2026
· 8 min read · Provenance · C2PA The label that falls off: where Content Credentials actually survive online The compliance plan for AI content rests on metadata travelling with the file. An audit of 516 posts found it survives to the viewer about thirty percent of the time. Read →10 Jul 2026
· 8 min read · Incidents · Brand safety Shipped without looking: what three years of public AI-creative fails have in common A third arm in a mayoral platform. Nineteen angry men. “KODAK ROMTOE” on a billboard. The catalogue of AI marketing fails keeps growing — and it is one story, told over and over. Read →10 Jul 2026
· 7 min read · Regulation · EU AI Act The EU’s AI-content labeling rules arrive on 2 August. Here’s what marketing teams actually have to do. In three weeks, disclosing realistic AI-generated imagery stops being a courtesy in the EU and becomes a legal obligation — one that lands on the advertiser, not the AI vendor. Read →