BitGo CEO Mike Belshe Challenges Anthropic to Prove Claude's Security with 100 BTC Test
Key Takeaways
- •BitGo CEO Mike Belshe has proposed a real-world security test in which 100 Bitcoin would be placed in a wallet to evaluate Anthropic's AI assistant, Claude.
- •The challenge is intended to move beyond theoretical security claims by testing Claude's robustness under realistic cybersecurity conditions.
- •Belshe's proposition is notable given BitGo's established position as a major institutional digital asset custody provider serving exchanges, funds, and enterprises since 2013.
- •Anthropic, founded in 2021 with a focus on AI safety, has not publicly responded to or indicated whether it will accept the challenge.
- •The proposed test underscores a broader industry gap in standardized third-party frameworks for independently verifying the security claims of AI models.
BitGo CEO Mike Belshe has publicly challenged Anthropic to validate the security claims surrounding its AI assistant, Claude, by placing 100 BTC into a cryptocurrency wallet as a real-world test.
The proposed challenge is designed to assess whether Claude's security capabilities can withstand practical conditions rather than relying solely on theoretical claims or controlled demonstrations. With 100 Bitcoin at stake—worth millions of dollars at current market prices—the test has drawn attention from both the cryptocurrency and artificial intelligence communities. Belshe's proposition carries particular weight given BitGo's position as a leading institutional digital asset custody provider, processing a significant share of Bitcoin transactions and serving major crypto exchanges, funds, and enterprises since the company's founding in 2013.
A High-Stakes Security Experiment
According to Belshe, the wallet containing 100 BTC will serve as the benchmark for evaluating Claude's security. The challenge reflects growing interest in understanding how advanced AI systems perform when confronted with realistic cybersecurity scenarios—a pressing concern as the digital asset industry has suffered billions in losses from exchange hacks, bridge exploits, and wallet compromises in recent years.
As AI tools become more deeply integrated into financial services and digital asset management, questions surrounding their security, reliability, and resistance to exploitation are becoming increasingly critical. The test also arrives amid broader industry debate over how AI companies—including Anthropic, which was founded in 2021 with a stated mission focused on AI safety—substantiate claims about their models' robustness against adversarial attacks and manipulation.
However, Anthropic has not publicly indicated whether it will accept the challenge.
LATEST: BitGo CEO Mike Belshe challenges Anthropic to prove Claude's security claims for real, putting 100 $BTC in a wallet as the test. Will Claude be able to crack it? pic.twitter.com/6SW7mpjyan — Cointelegraph (@Cointelegraph) August 3, 2026
AI and Crypto Security in Focus
The BitGo–Anthropic challenge underscores the intersection of artificial intelligence and cryptocurrency security. As digital assets continue to appreciate in value, safeguarding wallets and financial infrastructure remains a top priority for the industry. The challenge also highlights a broader gap in standardized, third-party security verification frameworks for AI models—a gap that public stress tests like Belshe's could help address if adopted more widely.
Whether or not the challenge proceeds, it has sparked broader discussion about how AI security claims should be validated and whether public, high-value tests can help build trust in emerging technologies.
Investors and technology observers will be watching closely for Anthropic's response and any further developments surrounding the proposed 100 BTC security test.