Ethical AI CX: Consent as Communication (1/3)
Below is the content from the Ethical AI Workshop organised the Ethical Commerce Alliance, London Sep 26th 2023. [UPDATE] You can find some personal notes on the event in this link.
Why is Ethical AI important in ecommerce?
Brands and retailers have a responsibility beyond designing great products: purpose. Expressed with values and principles, brands have a fundamental role in our lives in terms of promoting sustainability, empowering women, reducing inequalities, protecting our privacy, and many other human rights and meaningful purposes which our governments and the media slowly (and sloppily) seem to embrace.
Brands, however, need to be believed. Authenticity is key. We’re no fools. Brands need to go beyond beautifully written claims into actions. Ethics and Trust are pivotal.
The proliferation of AI only increases the need to ensure technology and automated systems elevate brand values and express them with care and respect.
Some brands start applying Ethical AI with protectionism, with #brandsafety as a core objective. Others apply Ethics in CX to connect with customers in more intimate ways, seeking new personalised experiences where customers always feel understood and in control, and are happy to share their personal data.
How do you design for Ethical AI?
Inspired by Reid Blackman in his book “Ethical Machines”, we decided to use the Big Three ethical layers framework:
- Explainability: Design for transparency and understanding.
- Privacy: Design for privacy and trust.
- Bias: Design for fairness.
3 Ethical AI CX Use Cases in Ecommerce that respect your customer’s privacy
This post will only cover the first:
- Consent as Communication
- Responsible Personalisation — Next post (2/3)
- Personal Shopper AI — Following post (3/3)
Consent as Communication
As our partner Rhiannon Hanger from Profound so brilliantly illustrated with visual examples in the workshop, why does consent need to be such a frustrating online experience full of repetitive, annoying banners and prompts we can’t understand?
Instead of devaluing our customers attention, why don’t retailers turn consent into an experience that evokes trust?
Instead of assuming customers are willing to accept consent as soon as the enter the ecommerce site [Imagine for a second how you would feel if the same would happen as you enter a physical store], why don’t retailers focus on demonstrating to their customers how they benefit from sharing their personal data?
Opt-in should be a conscious decision.
Dear retailers, always go for opt-out by default.
Earn your customer’s data (and trust).
Let’s look at three examples, ranking from Low to High Trust:
Low Trust: Customers are forced to click on “learn more”, “Reject All” is a 2nd click. Not only annoys customers, but also is unlawful.
Mid Trust: “Reject All” is not only a first layer click, but also selected by default. Compliant with privacy regulation, less annoying; yet, still remains opaque in its value proposition and degree of explainability.
High Trust: Consent as a conversation: Opt-out by Default + Progressive contextually informed consent.
- No cookie banner or any other pop up window is displayed when first time visitors enter the site.
- Asks for consent at the right time and context in terms of benefits.
- Informs when a personal preference is inferred (“Want to add Gluten Free to your preferences”), or a cart saved for future sessions “Before you go, do you want to keep your cart for next visit?”
As the above examples illustrate, Privacy, Explainability and Bias are key ethical principles that can be expressed in the way personal data is collected, processed and activated in CX.
Consent should not only be explicitly informed, but most importantly earned and its value explained.
Follow us to be informed when the upcoming posts on Ethical AI CX are published.
