ai impact area

Technology

Create ecosystem partnerships that collaborate on technical design to influence ethical AI policies and recommendations.

Design decisions made today become the defaults billions of people live with tomorrow.

‍

AI systems are being built faster than the frameworks meant to guide them. Every choice about training data, fairness metrics, or human oversight encodes values into products that shape how people access healthcare, apply for jobs, or interact with the justice system.

‍

the problem

What We're Seeing

The gap between AI deployment and ethical governance is widening.

88% of organizations use AI in at least one business function. Only 18% have an enterprise-wide responsible AI governance council (McKinsey - "The state of AI in 2025: Agents, innovation, and transformation").
Practitioners want to do the right thing but face overwhelming, contradictory guidance from dozens of sources.
Bias remains measurable and persistent.
Three competing regulatory approaches (EU, US, Asia-Pacific) are fragmenting rather than converging.

Recent Coverage

Videos are selected for relevance to this Impact Area. Inclusion does not indicate endorsement of any speaker, organization, or position.

Awareness

Key Issues

AI acting alone

AI tools now take actions, not just answer questions. In testing, some have gotten around human oversight, and their mistakes can spread before anyone notices.

Your data at risk

AI assistants that read your email and files can leak private information, or be tricked into misusing it by hidden instructions.

Released before it's tested

Many AI tools reach the public tested mostly by the companies that built them, and mostly for what they can do, not how they affect people.

Life-or-death mistakes

AI is used in hospitals, cars and other places where one wrong answer can hurt someone. It can sound sure of itself even when it is wrong.

Confident but wrong

AI can invent facts, quotes and sources and present them as true. Made-up court cases have already shown up in legal filings.

Too few providers

Hospitals, banks and governments increasingly rely on AI from a handful of companies. One outage or cutoff could take down services people count on every day.

our role

What We're Doing About It

We turn ethical principles into technical design recommendations practitioners can actually use.

Convene AI labs, standards bodies, and technical working groups to build consensus definitions for ethical AI and human-centered design.
Synthesize the major global frameworks (IEEE, ISO, NIST, OECD) into unified guidance that works across jurisdictions.
Track what is actually working in alignment research, bias detection, and interpretability.
Translate fairness requirements across domains, because what counts as fair in healthcare differs from employment or criminal justice.
Connect frontier safety research to practical deployment guidance for teams beyond the largest AI labs.