Page 10 - The Indian EYE 081823

The Indian Eye
OPINION AUGUST 18, 2023 10

US Voluntary Code of Conduct on AI

and Implications for Military Use

AKSHAT UPADHYAY Lawmakers in the US have been attempting to rein in the exponential developments in the AI field for some time now, since not much is known
about the real impact of the technology on a longer-term basis (File photo)
even technology companies
including Microsoft, OpenAI,
SAnthropic and Meta, with ma- public use. Bio, chemical and ra- • Publicly report model or system 2026 banning the use of AI in auto-
capabilities including discussions
mated weapons of war.
diological risks and ways in which
jor artificial intelligence (AI) prod- barriers to entry can be lowered for of societal risks; The discussion at the UNSC can
ucts made voluntary commitments weapons development and design • Accord priority to research on soci- be seen as elevating the focus from
regarding the regulation of AI at an are some of the top priorities. The etal risks posed by AI systems; and shorter term AI threat of disinfor-
event held in the White House on 21 effect on systems which have inter- • Develop and deploy frontier AI mation and propaganda in a bilateral
July 2023. These eight commitments actions and the ability to control systems to help address society’s context between governments and
are based on three guiding principles physical systems needs to be evalu- greatest challenges. Big Tech companies to a larger, glob-
of safety, security and trust. Areas ated apart from societal risks such The eight commitments of US’s al focus on advancements in AI and
and domains which are presumably as bias and discrimination; Big Tech companies come a few days the need to follow certain common
impacted by AI have been covered by • Information sharing amongst com- after the United Nations Security standards, which are transparent,
the code of conduct. While these are panies and governments. This is Council (UNSC) for the first time respect privacy of individuals whose
non-binding, unenforceable and vol- going to be challenging since the convened a session on the threat data is ‘scraped’ on a massive scale,
untary, they may form the basis for a entire model is based on secrecy posed by AI to global peace and se- and ensure robust cybersecurity.
future Executive Order on AI, which and competition; curity. The UN Secretary General Lawmakers in the US have been
will become critical given the increas- • Invest in cybersecurity and safe- (UNSG) proposed the setting up attempting to rein in the exponential
ing military use of AI. guards to protect unreleased and of a global AI watchdog comprising developments in the AI field for some
THE VOLUNTARY AI proprietary model weights; experts in the field who would share time now, since not much is known
COMMITMENTS ARE THE • Incentivize third party discovery their expertise with governments and about the real impact of the technol-
FOLLOWING: and reporting of issues and vulner- administrative agencies. The UNSG ogy on a longer-term basis. The re-
• Red-teaming (internal and exter- abilities; also added that UN must come up actions to the so-called danger of AI
nal) products to be released for • Watermarking AI generated content; with a legally binding agreement by Continued on next page... >>

www.TheIndianEYE.com