“The science is here,” UN Secretary-General António
Guterres said at the report launch. “We can no
longer say we did not know. What we do with it is now up to
all of us.”
The more AI advances without
shared rules, the less say governments and people will have
in the outcome, the UN chief said, adding “my
message to governments is simple: do not
wait.”
Aiming to build a shared
understanding and evidence at this critical juncture, the
Preliminary Report of the Independent International
Scientific Panel on AI: Evidence-based assessment of
opportunities, risks and impacts of AI was penned by the
first global, fully independent scientific body dedicated to
assessing its real impacts across economies and
societies.
Why it matters
Globally, over one
billion people now use conversational AI weekly, while
governments are making consequential decisions in the face
of great uncertainty with rapidly changing, often
conflicting sources of evidence and perspectives that do not
necessarily reflect local realities.
“Used well,
AI could be the most powerful engine for
development, speeding the world’s progress on everything
from health and hunger to learning and climate,”
the UN chief said, “but the panel is just as clear-eyed
about the harm artificial intelligence can
cause.”
Indeed, as the capabilities of AI continue
to grow, so do the stakes – the core challenge the panel
aims to address.
Better world or catastrophic
harm?
Advertisement – scroll to continue reading
Composed of 40 leading scientists and experts
from every region, the panel outlines AI trends and warns
that current safeguards cannot keep pace, said its co-chair
Yoshua Bengio.
“AI capabilities are
outpacing both scientific understanding and governments’
ability to adapt,” Mr. Bengio said. “With
growing evidence of deceptive AI behaviour, science
currently cannot guarantee that as capabilities continue to
increase, AI will not cause catastrophic harm, either on its
own or due to malicious users.”
To act effectively,
he said, global policymakers must understand these systems,
and the panel provides exactly that: a rigorous, shared
scientific foundation “to guide our collective way
forward”.
Key findings
Detecting breast
cancer earlier, accelerating vaccine development and
improving healthcare services are just a few ground-breaking
AI accomplishments, but limitations and challenges remain,
among them:
- AI adoption has accelerated broadly,
but unevenly across countries and sectors - Access and
usage vary widely, with adoption across the global south
lagging far behind the global north - Significant
differences in compute infrastructure and models exist
between advanced economies, reflecting existing
inequalities
Moreover, development is highly
concentrated, with recent estimates finding that the United
States accounts for 75 per cent of the computing power among
the world’s top 500 AI supercomputers, with China
accounting for 15 per cent, and that both countries’
companies develop almost all leading general-purpose
models.
Understanding risks
Understanding and
managing AI risks is essential, the report stated, with
panel co-chair Maria Ressa adding that risks to societies,
security and the human species are already “too
high”.
“The technology is transformative, but if
the world keeps moving along this trajectory, humanity will
fail to realise the gains it promises,” she
warned.
Here are some of the panel’s
warnings:
- There are no scientific guarantees that
AI agent systems will not violate instructions, and evidence
is accumulating of cases where they already are - AI
agent systems will soon complete tasks that currently take
human programmers days or weeks, but their deployment raises
urgent questions for labour markets, cybersecurity and the
controllability of future AI systems - Sycophantic AI
behaviour, where responses reinforce users’ existing
beliefs regardless of accuracy, has been linked to several
severe mental health incidents, including documented
deaths - Criminals and bad actors have been documented
using AI systems to assist in cyberattacks - Advanced
technical abilities may allow novice private actors to use
AI in malicious ways across a range of applications such as
fraud and disinformation - Reliable methods for
retaining control over highly autonomous AI systems are
lacking
Many of these harms fall
disproportionately on already disadvantaged populations, the
report stated.
Unlocking benefits, mitigating
risks
Minimising AI risks and benefiting from this
technological tool requires good governance.
Concrete
next steps to close current gaps exist, but each requires
sustained investment in Member State capacity to shape,
evaluate and deploy AI.
Realising these opportunities
safely requires dedicated investments and policies to
incentivise equitable access and reward innovation while
preventing the exploitation of the vulnerable.
‘AI
will not close divides itself’
Dozens of distinct
governance instruments that seek to embed ethics and human
rights in AI systems are already in use across
jurisdictions, but they are fragmented, concentrated among a
few corporations and rarely measure real-world
effectiveness, the report found.
Amandeep Gill,
Under-Secretary-General and Special Envoy for Digital and
Emerging Technologies, said the new report delivers shared
scientific language that decision makers can now use going
forward.
“AI will not close divides by itself,” he
said.
The benefits land where institutions, skills and
data already exist, and where they do not, the same
technology can displace workers, widen inequality and leave
communities dependent on systems built without them in mind,
he explained.
“Those realities are now on the
record, independently verified, and impossible to set
aside.”
The report’s findings will be presented to
governments at the inaugural UN Global Dialogue on AI
Governance, convening in Geneva on 6 and 7
July.
