Managing the Risks of AI: Setting Standards and Governance

Artificial intelligence (AI) has the potential to be as transformative as fire was when humans discovered it 1.5 million years ago. However, like fire, AI poses significant risks, some of which are existential. This article explores how we can manage these risks and the different paths we can take.

Setting Standards for AI

To manage the risks associated with AI, the first step is to set standards for its use. These standards can be voluntary or mandatory, set by either governments or industry groups. However, voluntary standards are not sufficient as only responsible companies tend to follow them. Therefore, we recommend that the standards be required, even if initially lower. Both governments and industry groups should lead the way, with a public-private partnership needed to achieve our goals.

Governing AI

There are two aspects of AI that need governance: its use and its creation. The level of governance should coincide with the level of risk, with some types of AI inherently dangerous and needing careful management, limitation, or restriction. However, we don’t know enough today to write all the regulations and rules required, so authoritative bodies need to be established to issue new rules as necessary. These bodies need to be quick and nimble, as existing industries and government bodies move too slowly.

Global Consensus and Existing Frameworks

The buy-in of all parties is critical for effective governance, but waiting for a global consensus is not necessary. Various global groups have been actively setting forth their views and publishing their output, with some having the force of law behind them. While nothing yet covers the full scope comprehensively, if you were to put them all together, you would be at a commendable starting point for this journey.

While the ride will be bumpy, humans will ultimately prevail in managing the risks associated with AI. We must harness AI for good and avoid burning down the world. By setting standards and establishing governance, we can ensure that AI is used responsibly and safely.