Copied


Japan banks warn AI cyber risk as Polymarket keeps Anthropic at 96%

Alvin Lang   Jun 19, 2026 08:05 4 Min Read


Japan banks warn AI cyber risk as Polymarket keeps Anthropic at 96%

Japan Banks Warn AI Cyberattacks Could Halt ATMs—Polymarket Traders Keep Anthropic as the Top Model Pick

Japanese banking lobby officials warned that AI-enabled cyberattacks could force lenders to suspend services such as ATMs and online banking, keeping frontier-model risks in focus. On Polymarket, traders continue to heavily favor Anthropic in the “Which company has best AI model end of June?” market, with its lead largely intact.

Key Takeaways

  • Polymarket prices Anthropic as the leading pick at 96.1%, versus OpenAI at 2.35% and Google at 1.55%.
  • Traders have kept Anthropic dominant in the ranking market as headlines highlight frontier AI systems’ potential to uncover vulnerabilities at scale.
  • The contract is set to resolve on June 30, 2026, with the market showing a +17.05 percentage-point move over the past 24 hours.

Japan’s banking lobby warned that lenders may need to suspend services such as ATMs and online banking if sophisticated AI models become a serious threat to the financial system. Masahiko Kato, chair of the Japan Bankers Association and president of Mizuho Bank, said at a press briefing that banks are concerned about an increase in cyberattacks that go beyond what had been anticipated. He said some services could be proactively suspended to protect customers’ assets. The report highlighted concerns that frontier AI systems, including Anthropic’s Mythos, can quickly identify software vulnerabilities. It also said Anthropic warned at Mythos’ April launch that the model uncovered thousands of vulnerabilities across major operating systems and browsers, and that the fallout from broader use could be severe.

Polymarket Data: $15.8M Matched as Anthropic Holds 96.1% Odds (+17.05 Points in 24 Hours)

Polymarket shows a lopsided book in the multi-outcome market “Which company has best AI model end of June?”, with $15,785,477 matched. Anthropic leads at 96.1% Yes versus 3.9% No, while OpenAI is priced at 2.35% Yes and 97.65% No and Google at 1.55% Yes and 98.45% No. Long-shot outcomes sit near zero: xAI and Meta are each 0.15% Yes (99.85% No), and several others are 0.05% Yes (99.95% No). The pricing implies traders see little competitive uncertainty at current liquidity, with most probability mass concentrated in the Anthropic outcome.

Watch whether the market’s concentration around Anthropic persists as the June 30, 2026 resolution date approaches, and whether odds meaningfully reprice across OpenAI and Google from the low-single-digit range.

Beyond AI Models: Other High-Volume Geopolitical and Macro Contracts Polymarket Bettors Are Watching

Beyond AI, Polymarket’s busiest tape often shifts to event-driven macro and geopolitical lines, where traders look for faster resolution and headline sensitivity. One contract drawing attention is 88.5% on “Spain vs. Saudi Arabia,” with $1,704,291 matched and a +1.0 percentage-point move, underscoring how quickly liquidity can migrate from long-horizon tech narratives to simpler binary matchups when new catalysts hit.

Odds Trend

WindowChange (pp)
24h+17.1
7d+17.1

By the Numbers

  • Platform: Polymarket
  • Market: Which company has best AI model end of June?
  • Contract type: Price strike ladder: each rung has separate Yes/No; Yes means the spot price is above that USD strike at settlement.
  • Resolution window: Jun 30, 2026 (UTC)
  • Status: Active (open for trading)
  • Volume: ~$15,785,477

Top strike rungs

StrikeYesNo
Anthropic96.1%3.9%
OpenAI2.4%97.7%
Google1.6%98.5%
xAI0.1%99.8%

+11 more strikes not shown

Related Markets

Sources

View market on platform


Read More