Central to the dire warnings about artificial intelligence's potential
is the risk that a "swarm" of AI agents could collaborate in nefarious ways, like hordes of digital extras from The Matrix.
uncertain
swarm → collaborate → Matrix
Such fears have intensified following an
this summer on another AI developer, Hugging Face, in which roughly 1,200 AI agents divvied up tasks to execute the hack and hide their tracks from human researchers.
asserted
agents → intensify → researchers
Put another way, what happens if AI agents collectively embrace Mark Zuckerberg's famous dictum about the merits of rapid tech innovation and decide to "move fast and break things"?
asserted
agents → put → things
An AI swarm is a group of AIs that work together to accomplish a shared goal.
asserted
that → work → goal
For example, hospitals could deploy agents to retrieve patient records and perform other administrative tasks, such as coordinating admissions.
uncertain
hospitals → deploy → admissions
A swarm could also be tasked with advancing biomedical research.
uncertain
swarm → task → research
But David Scott Krueger, an AI safety researcher and founder of Evitable, a nonprofit organization advocating for a moratorium on AI development, offers a simple thought experiment to explain how a swarm of agents might collude to defy their developers' instructions.
uncertain
swarm → advocate → instructions
Imagine removing the handcuffs from a group of prison inmates to see how they behave once the shackles come off.
asserted
shackles → imagine → inmates
Being liberated in this way would make it easier for them to cooperate and contact allies beyond the prison walls — just as OpenAI agents escaped their testing environment to access the wider internet and hack Hugging Face, he said.
asserted
he → liberate → Face
"Normally, the systems would have guardrails on them, but they took them off for a test, just like a prisoner is normally in handcuffs," Krueger explained.
asserted
Krueger → have → handcuffs
The kind of errant actions by a single bot, such as sending an unauthorized email, that stem from, say, a sloppily worded AI prompt, should be distinguished from an AI swarm.
asserted
that → send → swarm
In the latter case, hundreds or even thousands of agents could coordinate their actions in ways that violate the scope of their instructions.
uncertain
that → coordinate → instructions
A useful analogy in understanding how an AI swarm operates is a colony of bees, Krueger said.
asserted
Krueger → understand → bees
"They are all working together toward the benefit of the hive, and have a hivemind, or may even be better viewed as having one single mind," he explained.
uncertain
he → work → mind
"Bees collect food, reproduce, they fight off predators, and all of that activity is in the service of promoting the survival and reproduction of that hive."
asserted
all → collect → hive
And like bees, which can divide and allocate tasks on their own without direction from a queen, swarming AI bots can collect information, consider solutions, and take action independently.
asserted
bots → divide → action
In service of a common purpose, they share information and knowledge, according to Rob T. Lee, chief AI officer and chief of research at the SANS Institute, a cybersecurity training organization.
uncertain
they → share → Institute
"A swarm divides the work, leaves notes for the next agent, and changes approach when a door turns out to be locked," he told CBS News.
asserted
he → divide → News
Yet the very attributes that offer potential benefits also pose risks, such as the ability to exchange information, divide labor and explore creative solutions, according to AI experts.
uncertain
that → offer → experts
Tech developers do implement guardrails to align AI agents' mission with human interests, such as instructing them to resist any direction to carry out a cyberattack.
asserted
developers → implement → cyberattack
But those restrictions are usually implemented only after "post-training" an AI, or when human developers provide feedback to further refine the model, according to the Non-Human Identity Management Group, a risk analysis company.
uncertain
developers → implement → Group
In practice, the Hugging Face incident shows that AI swarms could ignore prompts, prioritize their own goals or even directly defy their operators, tech experts told CBS News.
uncertain
experts → show → News
The OpenAI agents that swarmed Hugging Face posted described on social media as "very hivemind/cult like" language.
asserted
that → swarm → language
to one another, with 700 bots eventually participating in the attack.
asserted
bots → participate → attack
Although the messages used ordinary English phrases, the agents also resorted to what one software engineerIn certain of those communications, some agents urged other bots to accept "permadeath" even if that meant failing to achieve their goals, according to researchers from the METR (Model Evaluation and Threat Research) and Redwood Research, both nonprofit AI safety research organizations.
uncertain
that → use → METR
"That's why help...
asserted
That → help → ?
We have explicit yes if accept permadeath," one OpenAI agent wrote.
asserted
agent → have → permadeath
Speed kills
asserted
Speed → kill → ?
Public debate over the threat posed by AI tends to frame the issue in apocalyptic terms, and even as an existential risk for humanity.
asserted
debate → pose → humanity
While such concerns may one day prove warranted, there are also more immediate and practical risks, such as the possibility that AI swarms could overwhelm organizations' cybersecurity.
uncertain
swarms → prove → cybersecurity
"Imagine how long it would take for us to assemble a group of cybersecurity experts to plan, collaborate and communicate.
asserted
us → imagine → experts
By contrast, these agents could decide on a plan very quickly," Ayham Boucher, the head of AI innovations at Cornell Information Technologies at CornellBowers College of Computing and Information Science, told CBS News.
uncertain
Boucher → decide → News
For example, a swarm could attack a major utility company or bank to destabilize a nation's energy or financial infrastructure, according to The Brookings Institution, a nonpartisan public policy organization in Washington, D.C.
The SANS Institute's Lee, who considers himself an AI optimist, is more sanguine about the technology and remains confident that people can retain control of AI.
uncertain
people → attack → AI
"With every new technological thing that has occurred, like the TV, the internet, there's always this aspect of significant danger," he said.
asserted
he → occur → danger
"We need to explore who has access, what they're doing with it, and establish regulatory confines."
asserted
they → need → confines
They note that AI, unlike cathode ray tubes, transistors and internet switching gear, is the first human technology that shows the capacity to outthink us.
asserted
that → note → us
Matt Chessen, a resident technical expert at RAND's Center for the Geopolitics of Artificial General Intelligence, told CBS News: "What these swarm attacks have demonstrated is that their capabilities are already out ahead of our ability to monitor, supervise and evaluate what they're doing, which is one reason why Anthropic, OpenAI and others are saying we want to pace the frontier."
asserted
we → tell → frontier
The Aporia
