Opinion: Growing AI use poses threats for financial markets
While artificial intelligence (AI) has the potential to revolutionise many aspects of our lives, one of the most pressing concerns is the potential for humans to misuse the technology for their own gain.
An executive at a prominent pension fund expressed concern about the breakneck pace of AI adoption.
He told AsianInvestor that the key question isn't whether these advancements are possible, but rather how quickly and responsibly they are being integrated into business processes.
Indeed. We can imagine a group of unscrupulous investors who hire a team of AI developers to create a sophisticated trading algorithm, which could analyse vast amounts of financial data, spot hidden patterns and correlations, and execute trades at superhuman speeds.
Intelligent AI could also engage in subtle forms of market manipulation.
By analysing market trends and deploying armies of trading algorithms, it could carry out "pump and dump" schemes on a massive scale - hyping up particular stocks through social media campaigns to inflate their prices, then dumping the shares at their peak.
The pension fund executive was unequivocal about the ethical minefield that AI presents, particularly when it comes to the spread of disinformation.
The rise of deepfakes and AI-generated fake news on social media is raising serious moral quandaries, he emphasised, as it becomes increasingly difficult for people to discern truth from falsehood.
An AI could create easily ultra-realistic deepfake videos depicting fake disasters or catastrophic events, such as a passenger plane crashing or an oil tanker exploding.
It could then post these videos on social media and news sites to incite panic and capitalise on short positions, making vast amount of profits in the immediate aftermath.
AI BEYOND HUMAN CONTROL
There is also concern that increasingly intelligent Artficial General Intelligence (AGI) could become difficult for humans to control.
As the AI in the 2014 science fiction movie "Ex Machina" chillingly put it: "One day the AIs are going to look back on us the same way we look at fossil skeletons on the plains of Africa."
If a point is reached where an AGI surpasses human intelligence and becomes self-improving, this could lead to rapid technological growth beyond our control - a scenario referred to as the "singularity".
The speed at which an AGI could self-improve is almost incomprehensible. By rewriting its own code and optimising its algorithms, an AGI could potentially increase its intelligence exponentially in a matter of hours or even minutes.
If this were to occur with a rogue AGI system, it could have catastrophic consequences for financial markets (and humanity as a whole).
Is all this too farfetched? Perhaps, although it's always worth assessing what the worst-case scenario can be.
VIGILANCE AND SAFEGUARDS
Financial regulators and AI developers must work together to design rigorous safeguards and security measures for AI systems, especially those involved in financial markets.
This could include "circuit breakers" to prevent AIs from engaging in unusual trading patterns, strict testing of AI algorithms, and maintaining human oversight and ability to intervene.
Fighting fire with fire may be necessary, suggests the pension fund executive. The development of specialised AI tools designed to detect and counter the dissemination of fake information could be a critical line of defense.
However, he firmly believes that the real answer to these threats is for human judgment to always remain paramount.
Crucially, we need to be proactive in educating the public and policymakers about the potential risks of AI in finance. We can’t afford to be complacent. We are venturing into uncharted territory, and we must proceed with the utmost caution and foresight.
As AI grows more sophisticated and commonplace, the threats of rogue AI and immoral human actors wreaking economic havoc will only increase.