Google announces an Advanced Technology External Advisory Council to consider ethical issues around AI like facial recognition and fairness in machine learning
Last June we announced Google's AI Principles, an ethical charter to guide the responsible development and use of AI in our research and products.
Context & Ripple Effects
The council is the third rung on a ladder Google has been climbing since staff unrest forced it to publish its seven-point AI Principles in June 2018, which ruled out AI for weapons and surveillance. It also follows the cross-industry ethics group Google formed with Amazon, Facebook, Microsoft, and IBM back in 2016, making this an attempt to add an external check on top of an internal charter.
The mandate — facial recognition and fairness in machine learning — targets exactly the areas where the Principles left day-to-day judgment calls unresolved. But the design of the board quickly drew fire: within days nearly half its members had resigned or were under fire over its composition, with critics arguing Google was treating AI ethics as a PR exercise.
First-order effects
- Council members face immediate public scrutiny of their affiliations and views, and several resign before the body ever meets — leaving Google without the external advice it just announced.
- Google's chief legal officer ends up leading an internal AI ethics council instead, shifting ethical review from outside advisers to executives inside the company.
Second-order effects
- With external legitimacy gone, Google converts the ethics apparatus into a product line: by mid-2020 it plans to sell AI ethics consulting services covering bias-spotting and project guidance, turning a reputational liability into revenue.
- Rivals watching the collapse get a cautionary template for their own advisory boards — the 2016 multi-company ethics group shows peers prefer collective, lower-scrutiny formats over individually exposed external councils.
Third-order effects
- If the pattern holds, corporate AI governance consolidates inside companies under legal and compliance leadership rather than independent outsiders, weakening the external-check model for emerging issues like facial recognition.
- Ethics review migrating in-house also sets up the longer shift toward monetized assurance — vendors grading their own homework while selling ethics guidance to customers, a structural conflict regulators may eventually have to address.
The trend: Corporate AI ethics is migrating from independent external advisers toward internal, legal-led councils — and increasingly toward paid governance services — leaving the external-check model largely untested.