Mohammad Nasravi

Chat with Mohammad on WhatsAppگفت‌وگو با محمد در واتساپتواصل مع محمد عبر واتساب

AI Daily: Meta’s AI edits your ad. You carry the risk

A robotic arm glowing cyan paints strokes onto an empty advertising frame on an easel in a dark navy studio, while a heavy iron weight on a chain waits on a stool in front.

Written by Mohammad Nasravi inنوشته شده توسط محمد نصراوی در

Meta’s updated Commercial Terms add an “AI Features” section: its AI tools “may make mistakes”, and the advertiser is solely responsible for the risks and the content. The page still says 30 October; Meta has told reporters 20 November. Check every Advantage+ creative toggle this week.

Meta: its AI edits, your liability

Meta’s preview of its new Commercial Terms says the updates take effect on 30 October 2026. Section 5, “AI Features”, says the features may make mistakes and that the business alone carries the risks (Meta Commercial Terms preview, checked 25 Oct). Adweek reports that Bookshop.org found an AI illustration added to its human-made ad after “an account update opted us in”, and that the agency Exverus pulled live ads after Meta added sitelinks, translated copy and built images from brand photos. Meta says each Advantage+ feature can be switched off in Ads Manager (Adweek, 22 Oct). On 23 October a Meta spokesperson told Reuters the date moves to 20 November; the preview page still showed 30 October that evening (PPC Land, 23 Oct).

My read: I throw away 60 to 70 percent of the takes the tools give me. An enhancement Meta switches on is a take nobody on your team rejected, running under your logo, and under these terms it is legally yours. This is the same pattern as ChatGPT Ads, which start rewriting by default on 29 October (my note from Wednesday). Treat the date on the page as binding, not the one in a spokesperson’s email.

Pennsylvania makes misleading AI ads a legal matter

Governor Josh Shapiro approved Senate Bill 806 on 22 October; it is now Act 99 of 2026 (Pennsylvania General Assembly, bill history). It requires a disclosure when AI-generated or AI-modified content in an ad could mislead people about the “nature, origin or authenticity” of what is sold, in the same medium as the ad (MediaPost, 8 Oct). The sponsor, Senator Nick Pisciottano, says it applies to anything shown in Pennsylvania, even if it was made out of state; it takes effect 60 days after signing (CBS News Pittsburgh, 22 Oct), around 21 December.

My read: The test is the product, not the tool. In the Popel knitted ad the people, the pool and the clouds are yarn; nobody is misled about a cookie. A photoreal milk pour that never happened is not. I have localised one hero film into several language versions for a regional rollout. Under a law like this, each cut is its own ad, and the disclosure has to travel with every one of them.

Canva Craft: an image model trained on templates

Canva released Craft on 22 October, its first image model built and trained in-house by Canva Research. It plans the composition first, then renders at up to 2K in more than 30 aspect ratios, renders text in 55 languages, and is live for all Canva users in Dream Lab and through the public API (Canva Newsroom, 22 Oct). The training data includes millions of human-made templates from Canva’s library. It published no language list and no price.

My read: Pack text is the hardest test of “text that lands where it should”. In the Popel film the one thing that is not yarn is the real printed pack, in every shot. I would not let any model redraw a pack’s lettering. Craft’s template training may suit social posts and posters; run your own brand fonts and your second language through it before it goes near a brief.

What I’d do today

Open Ads Manager, go campaign by campaign, and switch off every Advantage+ creative enhancement you have not approved by name. Then save a screenshot of each setting with the date. If the AI work is the point of the brief, I direct it frame by frame instead of leaving it to defaults.

Sources

خبرهای هوش مصنوعی را در تلگرام دنبال کنخلاصهٔ خبرها و برداشتِ من از کارِ واقعی، در کانالِ Nasravi.com در تلگرام.
عضویت در کانال تلگرام

Written by Mohammad Nasravi — AI creative director, in-house at a major FMCG group. Four years putting AI-made commercials past real brand teams, including Shirin Asal campaigns for Gulfood 2026 and Canton Fair 2025. More about the work · Start a project نوشتهٔ محمد نصراوی — کارگردان خلاق هوش مصنوعی، درون یک گروه بزرگ FMCG. چهار سال است کارِ ساخته‌شده با هوش مصنوعی را از فیلتر تیم‌های برند رد می‌کند؛ از جمله کمپین‌های شیرین عسل برای گلفود ۲۰۲۶ و کانتون فر ۲۰۲۵. دربارهٔ کارها · شروع پروژه بقلم محمد نصراوي — مدير إبداعي للذكاء الاصطناعي من داخل مجموعة FMCG كبرى. أربع سنوات من تمرير إعلانات مصنوعة بالذكاء الاصطناعي عبر فرق علامات حقيقية، منها حملات Shirin Asal في جلفود 2026 ومعرض كانتون 2025. المزيد عن العمل · ابدأ مشروعك