Page 64 - EE Times Europe Magazine | February 2020

62 EE|Times EUROPE

OPINION

What’s the Best Strategy

for AI Regulation?

By Sally Ward-Foxton

During CES, the White House issued a draft memo to U.S. govern-
ment agencies with guidance on creating regulations for the use of
artificial-intelligence technology in the private sector. The memo,
from the White ouse Office of Science and Technology Policy
(OSTP), was intended both to promote the trustworthy use of AI
technology and help avoid its overregulation. Preventing authoritarian use of AI
Are those two goals at odds depends on becoming a global leader in
The White ouse thinks not. Its position is that the way to become the technology. (Image: Scott Webb/Pexels)
a global leader in AI is to avoid unnecessary regulation, and the best
way to ensure ethical use of the technology is to be a global leader in it.
Indeed, the White ouse went as far as to warn Europe in strong terms not to overregulate. As the White House warns
“Europe and our allies should avoid heavy-handed innovation-killing models,” reads a
statement issued by the OSTP. “The best way to counter authoritarian uses of AI is to make sure Europe against overregulation,
America and our international partners remain the global hubs of innovation, shaping the evolu-
tion of technology in a manner consistent with our common values.” is ethical use of AI necessarily
Europe’s approach to AI regulation is not yet set in stone, but all the signs point toward strin- an innovation killer?
gent legislation — the opposite of the approach favored by the U.S.
European Commission president Ursula von der Leyen set the tone when she made the human
and ethical implications of AI a cornerstone of her pre-election manifesto, promising to put the public is essential to the development of a
forward legislation for a coordinated European approach. sustainable AI industry and, by extension, the
A recent report by Germany’s Data Ethics Commission recommended tough new rules for AI evolution of AI technology.
ethics, with strong measures taken against “ethically indefensible uses of data.” This was widely Public trust is not something to gamble
seen as an indication that any new EU rules on AI uses would be just as tough, given that a previ- with — once it’s lost, it’s gone forever — so
ous report from the Data Ethics Commission was the basis for the EU’s General Data Protection regulation is needed to maintain it. GDPR has
Regulation (GDPR). been widely copied, from apan to Brazil. If it
plays its hand wisely, Europe could once again
TO REGULATE OR NOT TO REGULATE? lead the world in this area.
It’s clear that regulation will play a key role in shaping the emerging landscape as global super- Proving that ethical use of AI is not neces-
powers jostle to gain expertise in this game-changing technology. But which strategy is better, sarily an innovation killer will be hard, but it
and what is the right balance between regulation and innovation is not worth compromising over. ■
Is regulation required to prevent the authoritarian use of AI — already an alarming trend for
applications such as facial recognition in some countries My gut says that the industry cannot Sally Ward-Foxton is a staff correspondent
be expected to police itself on issues as critical as civil liberties. Ultimately, earning the trust of at AspenCore.

IMAGES S UTTERSTOCK

Maintaining the public’s trust in AI technology is essential to the development of a sustainable AI industry. (Image: Andy Klly/Unsplash)

FEBRUARY 2020 | www.eetimes.eu