A principled approach to AI

We are at a juncture with artificial intelligence (AI) very similar to where we were with electricity more than a century ago. Sure, electricity was revolutionary with the capacity to reinvent—well—everything. But it could also burn down houses were it not for the circuit breaker.

Today, AI holds the promise of similarly revolutionary potential. It can improve transparency in government and better the lives of citizens worldwide. It augments our human potential and could launch the next industrial revolution. AI can be harnessed for good in practically every cause worth pursuing—from addressing climate change to improving government function.

AI might offer significant advantages but can also be weaponized when used in harmful ways. Deepfakes that spread misinformation and undermine democratic principles are just one of the many ways in which ill-intentioned entities can use AI to achieve undesirable end goals.

Whether it was electricity then or AI today, the take-home message is clear: all promising technology needs robust guardrails. And for ACTS, those guardrails rest in our principled approach to AI.

A principled foundation for ACTS

Working closely with governments and through public-private partnerships, we at ACTS have seen AI’s potential for good (and bad) up close.

It is why we anchor our work in a core set of guiding principles for responsible AI that governs all work at Microsoft. Under these umbrella guidelines, Microsoft is committed to ensuring that our AI systems are accountable and transparent, respect privacy, and operate securely.

Equally important, responsible AI at Microsoft is not an afterthought or just the cherry on top. We are committed to responsible AI from the outset, and we bake these principles into our design. Responsible AI is the structural framework with which we build, including ACTS.

ACTS walks the talk

ACTS goes further: using the Microsoft responsible AI principles as a foundation, we have built sector-specific guardrails that ensure transparency and trust in resultant systems and processes. We think ten steps ahead and evaluate all foreseeable uses for solutions so we can ensure not just responsible production but also responsible implementation.

ACTS thoroughly vets the proofs of concept that we build for governments, and we share our processes and recommendations for potential use. Such knowledge-sharing and internal reviews ensure that we have run through our own internal checklist for responsible development of AI solutions.

And a principled approach means ensuring that the solutions will be used for good. ACTS has not hesitated to withhold our technology from entities who might use it for negative purposes, an important cornerstone of the responsible AI guardrails.

Building on trust

Closely connected with the concept of a principled approach is trust. ACTS works hard to earn it and ensure that we never  take for granted the trust our partners place in us.

There’s a reason ACTS sits under Microsoft’s Corporate, External, and Legal Affairs (CELA) umbrella: CELA, too, is about building trust with our customers. ACTS delivers transparency solutions to governments and other public-facing entities around the world, which overlaps with CELA’s work on the cutting edge of business and regulatory issues. We mobilize the power of data and technology to assist governments  in accelerating transparency.

Our public sector customers can trust that we will handle their data securely and only in certain specific agreed-upon ways, respecting data sovereignty requirements. ACTS already checks off all the stringent privacy guidelines in our responsible AI framework, and we also ensure compliance with local data-handling laws.

Trust features again in how we use technologies like generative AI. This headline-grabbing technology has very real positive implications for how we can conduct business. Generative AI will allow us to, among other capabilities, embed new functionalities like natural language queries in ACTS. Such powerful enhancements will mean citizens and government employees can use chat boxes and easily understandable language to ask questions and get the information they need to conduct business smoothly. It’s about enhancing workflows and making government function faster and better.

Building a bright AI-driven future

We recognize that we need to build on a principled approach to AI and a strong foundation of trust.

Part of our commitment to treating data with care is embedded in our principled AI approach. We are on the cusp of a meteoric growth phase with respect to AI adoption. There is much to be excited about with respect to the transformative power of this technology. With the appropriate guardrails in place, ACTS is committed to building valuable AI-driven solutions for the public sector that will make government more transparent, more accountable, and better equipped to improve citizens’ lives today—and tomorrow.