You are here

German firm draws up seven commandments for ethical AI

BT_20180920_WEESAP20_3566834.jpg
SAP said its guiding principles aim to steer the development and use of its AI software so the world runs better.

Frankfurt

GERMAN business software giant SAP has published an ethics code to govern its research into artificial intelligence (AI), in a bid to prevent the technology infringing on people's rights, displacing workers or inheriting biases from its human designers.

The company said in a statement: "SAP designed these guiding principles to steer the development and deployment of our AI software to help the world run better and improve people's lives."

A group comprising engineers, theologians, political scientists and bio-ethicists put together the list of seven guidelines, ranging from respect for United Nations human rights principles to data protection.

sentifi.com

Market voices on:

Following in the footsteps of better-known US tech giants like Google and Microsoft, SAP - which makes software for some 400,000 firms world-wide to manage data on customer relationships, payrolls or inventory - is wrestling with how to impose values on AI's systems of algorithms designed to ape certain aspects of human intelligence.

Public figures such as entrepreneur Elon Musk and physicist Stephen Hawking have warned of the dangers of out-of-control AI.

Mr Musk said at a technology conference this year: "The danger of AI is much greater than the danger of nuclear warheads by a lot."

But rather than scenarios like those depicted in modern movie classics The Terminator or The Matrix - that of computer programmes going rogue against their human creators - SAP's concerns are closer to home.

Its seven-point list covers contemporary concerns like transparency, safety and privacy. And one item on the list specifically warns of "a risk of causing discrimination or of unjustly impacting under-represented groups" if AI inherits biases from human programmers - a danger SAP hopes to avert with more diverse development teams as well as "technical methods".

Meanwhile, it is the potential effect on jobs that is driving many warnings about the technology.

In an interview published on the SAP website, chief financial officer Luka Mucic acknowledged that some people's jobs would disappear as machines take on more work.

Nevertheless, "the changes brought by AI will most likely take shape just as other major technological developments throughout history, where new types of jobs are created to replace lost ones, he added.

Powerful German union IG Metall also chimed in on AI on Tuesday. The union's co-president Christiane Benner, calling in a statement for firms to consult early with employee representatives before implementing the technology, said: "Our answer to the challenges of AI is critical intelligence. AI applications should unburden people and create free space for higher-value activities."

Not to be outdone by business and workers, Germany's federal government this month convened a data ethics commission of its own, with ethical principles for AI among a range of fields to be drawn up within a year. AFP