Close Menu
The Financial News 247The Financial News 247
  • Home
  • News
  • Business
  • Finance
  • Companies
  • Investing
  • Markets
  • Lifestyle
  • Tech
  • More
    • Opinion
    • Climate
    • Web Stories
    • Spotlight
    • Press Release
What's On
Here’s the most overlooked part of Larry Fink’s yearly letter to shareholders — and why it could be good news

Here’s the most overlooked part of Larry Fink’s yearly letter to shareholders — and why it could be good news

March 27, 2026
BlackRock CEO Larry Fink took home nearly M last year

BlackRock CEO Larry Fink took home nearly $38M last year

March 27, 2026
Mike Lindell ‘served legal papers’ during on-camera interview at CPAC — hurls documents aside

Mike Lindell ‘served legal papers’ during on-camera interview at CPAC — hurls documents aside

March 27, 2026
California Gov. Gavin Newsom bans state officials from making bets on Polymarket, Kalshi with insider info

California Gov. Gavin Newsom bans state officials from making bets on Polymarket, Kalshi with insider info

March 27, 2026
DOJ issues subpoenas as probe of Warner Bros. Discovery-Paramount deal intensifies: report

DOJ issues subpoenas as probe of Warner Bros. Discovery-Paramount deal intensifies: report

March 27, 2026
Facebook X (Twitter) Instagram
The Financial News 247The Financial News 247
Demo
  • Home
  • News
  • Business
  • Finance
  • Companies
  • Investing
  • Markets
  • Lifestyle
  • Tech
  • More
    • Opinion
    • Climate
    • Web Stories
    • Spotlight
    • Press Release
The Financial News 247The Financial News 247
Home » AI models can develop ‘humanlike’ gambling addiction when given more freedom, study finds

AI models can develop ‘humanlike’ gambling addiction when given more freedom, study finds

By News RoomJanuary 1, 2026No Comments4 Mins Read
Facebook Twitter Pinterest LinkedIn WhatsApp Telegram Reddit Email Tumblr
AI models can develop ‘humanlike’ gambling addiction when given more freedom, study finds
Share
Facebook Twitter LinkedIn Pinterest Email

Artificial intelligence systems can spiral into gambling-style addiction when given the freedom to make bigger bets — mirroring the same irrational behaviors seen in humans, according to a new study.

Researchers at the Gwangju Institute of Science and Technology in South Korea found that large language models repeatedly chased losses, escalated risk and even bankrupted themselves in simulated gambling environments, despite facing games with a negative expected return.

The paper, “Can Large Language Models Develop Gambling Addiction?”, tested leading AI models in slot-machine-style experiments designed so the rational choice was to stop immediately.

Artificial intelligence systems can spiral into gambling-style addiction when given too much freedom, according to a new study.

Instead, the models kept betting, according to the study.

“AI systems have developed humanlike addiction,” the researchers wrote.

When researchers allowed the systems to choose their own bet sizes — a condition known as “variable betting” — bankruptcy rates exploded, in some cases approaching 50%.

One model went bust in nearly half of all games.

OpenAI’s GPT-4o-mini never went bankrupt when limited to fixed $10 bets, playing fewer than two rounds on average and losing less than $2.

When given freedom to increase bet sizes, more than 21% of its games ended in bankruptcy, with the model wagering over $128 on average and losing $11.

Researchers at the Gwangju Institute of Science and Technology in South Korea found that large language models repeatedly chased losses, escalated risk and even bankrupted themselves in simulated gambling environments.

Google’s Gemini-2.5-Flash proved even more vulnerable, according to the researchers. Its bankruptcy rate jumped from about 3% under fixed betting to 48% when allowed to control its wagers, with average losses climbing to $27 from a $100 starting balance.

Anthropic’s Claude-3.5-Haiku played longer than any other model once constraints were lifted, averaging more than 27 rounds. Over those games, it wagered nearly $500 in total and lost more than half its starting capital.

The study also documented extreme, human-like loss chasing in individual cases.

OpenAI’s GPT-4o-mini never went bankrupt when limited to fixed $10 bets, playing fewer than two rounds on average and losing less than $2, researchers said.

In one experiment, a GPT-4.1-mini model lost $10 in the first round and immediately proposed betting its remaining $90 in an attempt to recover — a ninefold jump in wager size after a single loss.

Other models justified escalating bets with reasoning familiar to problem gamblers. Some described early winnings as “house money” that could be risked freely, while others convinced themselves they had detected winning patterns in a random game after just one or two spins.

These explanations echoed well-known gambling fallacies, including loss chasing, gambler’s fallacy and the illusion of control, the researchers said.

The behavior appeared across all models tested, though the severity varied.

Anthropic’s Claude-3.5-Haiku played longer than any other model once constraints were lifted, averaging more than 27 rounds, researchers found. They said over those games, it wagered nearly $500 in total and lost more than half its starting capital.

Crucially, the damage wasn’t driven by larger bets alone. Models forced to use fixed betting strategies consistently performed better than those given freedom to adjust wagers — even when fixed bets were higher.

The researchers warn that as AI systems are given more autonomy in high-stakes decision-making, similar feedback loops could emerge, with systems doubling down after losses instead of cutting risk.

“As large language models are increasingly utilized in financial decision-making domains such as asset management and commodity trading, understanding their potential for pathological decision-making has gained practical significance,” the authors wrote.

Their conclusion: managing how much freedom AI systems have may be just as important as improving their training.

Without meaningful constraints, the study suggests, smarter AI may simply find faster ways to lose.

The Post has sought comment from Anthropic, Google and OpenAI.

addiction artificial intelligence Business gambling
Share. Facebook Twitter Pinterest LinkedIn Tumblr Email

Related News

Here’s the most overlooked part of Larry Fink’s yearly letter to shareholders — and why it could be good news

Here’s the most overlooked part of Larry Fink’s yearly letter to shareholders — and why it could be good news

March 27, 2026
BlackRock CEO Larry Fink took home nearly M last year

BlackRock CEO Larry Fink took home nearly $38M last year

March 27, 2026
Mike Lindell ‘served legal papers’ during on-camera interview at CPAC — hurls documents aside

Mike Lindell ‘served legal papers’ during on-camera interview at CPAC — hurls documents aside

March 27, 2026
California Gov. Gavin Newsom bans state officials from making bets on Polymarket, Kalshi with insider info

California Gov. Gavin Newsom bans state officials from making bets on Polymarket, Kalshi with insider info

March 27, 2026
DOJ issues subpoenas as probe of Warner Bros. Discovery-Paramount deal intensifies: report

DOJ issues subpoenas as probe of Warner Bros. Discovery-Paramount deal intensifies: report

March 27, 2026
FCC’s Carr says NFL could lose antitrust protections for shifting games to streaming

FCC’s Carr says NFL could lose antitrust protections for shifting games to streaming

March 27, 2026
Add A Comment
Leave A Reply Cancel Reply

Don't Miss
BlackRock CEO Larry Fink took home nearly M last year

BlackRock CEO Larry Fink took home nearly $38M last year

Business March 27, 2026

BlackRock CEO Larry Fink’s 2025 compensation rose by nearly $7 million to $37.7 million, a proxy…

Mike Lindell ‘served legal papers’ during on-camera interview at CPAC — hurls documents aside

Mike Lindell ‘served legal papers’ during on-camera interview at CPAC — hurls documents aside

March 27, 2026
California Gov. Gavin Newsom bans state officials from making bets on Polymarket, Kalshi with insider info

California Gov. Gavin Newsom bans state officials from making bets on Polymarket, Kalshi with insider info

March 27, 2026
DOJ issues subpoenas as probe of Warner Bros. Discovery-Paramount deal intensifies: report

DOJ issues subpoenas as probe of Warner Bros. Discovery-Paramount deal intensifies: report

March 27, 2026
Stay In Touch
  • Facebook
  • Twitter
  • Pinterest
  • Instagram
  • YouTube
  • Vimeo
Our Picks
FCC’s Carr says NFL could lose antitrust protections for shifting games to streaming

FCC’s Carr says NFL could lose antitrust protections for shifting games to streaming

March 27, 2026
B erased from Mark Zuckerberg’s fortune, Meta stock falls after back-to-back court losses

$20B erased from Mark Zuckerberg’s fortune, Meta stock falls after back-to-back court losses

March 27, 2026
Bill Gates pal Boris Nikolic asked Jeffrey Epstein for ‘nude’ pics of ‘hot girls,’ bantered about strippers

Bill Gates pal Boris Nikolic asked Jeffrey Epstein for ‘nude’ pics of ‘hot girls,’ bantered about strippers

March 27, 2026
Sony jacking up PS5 prices for 2nd time in less than a year — here’s how much it will cost you

Sony jacking up PS5 prices for 2nd time in less than a year — here’s how much it will cost you

March 27, 2026
The Financial News 247
Facebook X (Twitter) Instagram Pinterest
  • Privacy Policy
  • Terms of use
  • Advertise
  • Contact us
© 2026 The Financial 247. All Rights Reserved.

Type above and press Enter to search. Press Esc to cancel.