We need your opinion on AI manipulation

📅 Deadline to submit your input: 29 August 2026

𝐀𝐈-𝐞𝐧𝐡𝐚𝐧𝐜𝐞𝐝 𝐧𝐮𝐝𝐠𝐢𝐧𝐠: 𝐭𝐡𝐞 𝐦𝐨𝐬𝐭 𝐢𝐦𝐩𝐨𝐫𝐭𝐚𝐧𝐭 𝐀𝐈 𝐬𝐭𝐚𝐧𝐝𝐚𝐫𝐝 𝐲𝐨𝐮’𝐯𝐞 𝐧𝐞𝐯𝐞𝐫 𝐡𝐞𝐚𝐫𝐝 𝐨𝐟?

‘Nudging’ consumers to click on a particular link, buy a slightly more expensive product, or linger on a more extreme fb post, does not force or coerce them.

But it’s a well-developed technique to systematically and predictably change decisions and behaviors.

We can see physical nudges (putting candy in the check-out line at the grocery store) and digital nudges (“only 4 seats left!”), if we look closely enough.

But not necessarily AI-enhanced nudges:

• nudges that not even a business has decided on.

• that aren’t auditable.

• where no one can explain the decision logic.

• and where the actual results might be so unpredictable than properly conducted risk assessments, havent handled them.

ISO and CEN/CENELEC need public input on a new standard on exactly this: ISO/IEC DIS 25029, AI-enhanced nudging.

If you're in Norway, sign in to comment here: prNS-EN ISO/IEC 25029:2026 | Høring

All input is needed.

• If you are a provider and have opinions about your existing risk management - send them!

• If you are thinking about consumers, children, or workers - the standard explicitly covers all of these groups.

• If you are an ethicist - I think you will LOVE this standard.

More info?

🔗 Enrico Panai wrote a great guest post on Adam Leon Smith's substack The Standard that watches the watcher - by Enrico PANAI

🎙️ Cansu Canca talked about just how surrounded we are by AI nudges on every single digital platform The Puzzle of AI Ethics with Dr. Cansu Canca - Curiosity Weekly | Podcast on Spotify

📑 A recent experiment looking at Amazon's nudges AI nudging and decision quality: Evidence from randomized experiments in online recommendation setting - ScienceDirect

We on Standard Norway's AI committee haven't dug deep into this standard, because of the huge push for the EU AI Act harmonized standards. So please send us your thoughts!