Meta’s new terms leave advertisers liable for its AI ad edits
From 30 October, businesses carry the risk of whatever Meta’s AI features add to or change in their ads, even features Meta turns on itself, Adweek reports.
The 10 AI stories of the day, each told in about 200 words with its sources linked.
Published 06:30 UTC, 10 stories
From 30 October, businesses carry the risk of whatever Meta’s AI features add to or change in their ads, even features Meta turns on itself, Adweek reports.
Canva’s first in-house image model plans a layout before drawing it, and entered DesignArena’s crowdsourced leaderboard in fourth place, Canva says.
Investors from the UAE and Coatue lead the round, CNBC reports; a month ago, reports put the target at $1.4 trillion.
Each agent gets a profile inside the customer’s account and can be capped at a set amount; agents that break Walmart’s new rules get blocked.
On its earnings call, Meta also put ads run with creators at a $15 billion yearly run rate, up more than 97 per cent on a year earlier.
From 4 December, image extraction is on by default in new campaigns, even where an advertiser switched dynamic images off for the whole account.
In a study published in Science, MAI-DxO reached the right diagnosis in 83 per cent of complex cases, against 36 per cent for doctors using references.
Owners upload their colours, fonts, brand kit or website once, and Gemini applies that style whenever it draws marketing images with Nano Banana 2.1.
The standard for reusable agent instructions joins MCP at the Linux Foundation group, which took in projects from OpenAI, Hugging Face and Alibaba the same day.
The agent asks to log in, you approve it, and 1Password fills in the credentials itself; agents from Meta, Anthropic and OpenAI are among those supported.
adtestbench counts visits with Google Analytics cookies. Privacy and cookies