The 1985 deal has less than 8,000 words and contains just 22 articles,
mostly dedicated to tariffs, agricultural restrictions, import licensing
and rules of origin – what
Harvard economist Dani Rodrik calls conventional trade topics
.
While these issues are also covered in the US-Singapore deal, most of its
20 chapters and 70,000 or so words deal with other topics such as
anti-competitive business conduct, e-commerce, intellectual property,
investment rules, labour rights and the environment.
94% in just 26 seconds
AI has already proved its worth in the comparable field of law.
A two-month test pitting 20 lawyers against LawGeek’s AI showed that
humans were no match for a robot
in spotting risks within the legal documentation for non-disclosure
agreements – deals meant to protect confidential information such as new
manufacturing processes and marketing schemes.
In terms of accuracy, the lawyers scored an average of 85%, compared to the
robot’s 94%. But more impressive was the fact that AI was able to review
the five agreements in just 26 seconds, while the lawyers took an average
of 92 minutes.
“The accuracy and speed of AI in reviewing legal documents shows clearly
the game-changing potential of technology for trade negotiations,” Mr.
Onguglo said.
Restoring confidence
Just as important, the new technology could also make trade negotiations
more transparent for everyone, including the public, said Gabriel Petrus,
ICC Brazil’s executive director.
“There’s no better way to restore confidence in the global rules-based
trade system,” Mr. Petrus
Tweeted
on 3 June, referring to the anti-globalization tide that has swept many
parts of the globe as more workers and consumers have become convinced that
free trade has dealt them a bad hand.
In April 2018,
UNCTAD and ICC Brazil signed a partnership
to work together to use cutting edge technologies to boost and improve
trade.
ICC Brazil invited UNCTAD to join the artificial intelligence project
because the organization has expertise in the challenges that developing
countries face in trade negotiations.
UNCTAD took part in a “design thinking” session hosted by IBM Brazil on 30
May as part of ICC Brazil’s Intelligent Tech and Trade Initiative
. The session identified the main challenges of trade negotiations and
defined priority areas where AI could help.
The next step is for IBM Brazil to build a prototype of the tool and test
it on the trade negotiations between South American trade bloc Mercosur and
Canada. The results are planned to be presented at the
World Trade Organization’s Public Forum in October
.