Chamath Palihapitiya says AI could hack bitcoin and send it to zero

What happened

Billionaire investor Chamath Palihapitiya has said bitcoin's long-run value could collapse to nothing, arguing that artificial intelligence may eventually break the cryptography that secures the network. He made the remarks on Episode 304 of Peter Diamandis' Moonshots podcast, released earlier this week, according to Bitcoin.com News. Coinpedia and ZyCrypto both reported the comments as an interview.

Asked whether bitcoin was finished, Palihapitiya replied: "I just think it's capped, and then I think eventually it gets hacked, and then it goes to zero." His case rests on three points: the cryptocurrency's limited fungibility, what he describes as insufficient financial privacy, and the risk that increasingly capable AI systems could undermine the network's security.

The founder and CEO of Social Capital sketched a hypothetical attack in which an AI system cracks Bitcoin's defences or reaches the dormant wallets attributed to creator Satoshi Nakamoto. An attacker could take the coins, open a short position and then publicise the breach, he suggested, profiting from the resulting panic as well as from the theft itself. He also called a coordinated shift to quantum-resistant cryptography "basically impossible," pointing to the community's contentious governance and divided attention.

Palihapitiya added that bitcoin's transparent ledger makes it impractical for central banks to hold as a reserve asset, since sovereign institutions require confidentiality. Without that class of buyer, he argued, the asset faces a ceiling. He cited reports of kidnappings and home invasions involving crypto holders in France as evidence of the risks that come with publicly traceable holdings.

Why it matters

The comments carry weight mainly because of their source. Palihapitiya was among the most visible mainstream advocates of bitcoin, and his shift gives ammunition to critics who argue the asset's security model has not been tested against future computing power. The reports stress that the threats he describes are hypothetical rather than demonstrated: SHA-256 hashing, which secures mining, and the signature schemes that protect wallets are separate mechanisms, and Bitcoin has not suffered a catastrophic protocol breach of the kind he imagines. Still, the intervention puts a prominent voice behind questions about privacy and long-term cryptographic resilience that developers already debate.

What the data shows

The reported figures trace both his enthusiasm and his exits: entry prices of $80 to $100 between 2012 and 2013, a $1.6 million land purchase funded with about 2,739 BTC in 2014, a $200,000 target set in January 2021, and a self-described multi-billion-dollar opportunity cost from selling early. ZyCrypto reported that he now views spot bitcoin ETFs as a milestone that widened access, letting investors gain exposure without managing wallets or private keys.

Background

Palihapitiya's earlier positions were very different. He accumulated bitcoin around 2012 and 2013 at roughly $80 to $100 per coin, recommended allocating 1% of personal wealth to it as insurance against financial instability, and by 2017 was outlining targets of $100,000 and later $1 million. In January 2021 he projected $200,000 within five to 10 years. His holdings had already changed: in 2014 he reportedly used about 2,739 BTC, then worth $1.6 million, to buy land near Lake Tahoe, and on the All-In podcast in November 2024 he described earlier sales as a "$3 or $4 billion mistake… growing now." He now says he holds no bitcoin directly, having swapped long-term exposure for equity in Digital Currency Group to reduce custody risk. Bitcoin.com News noted a pattern of revised positions, including selling Tesla shares while publicly praising the company and later acknowledging flaws in the SPAC structure he popularised. He has said he "reserves the right to change my mind."

What to watch

The practical questions are whether AI or quantum research produces a credible attack on Bitcoin's signature or hashing algorithms, and whether the developer community can agree on and ship post-quantum protections if one emerges. Neither has happened yet. Palihapitiya's forecast is a prediction about future technology, not a description of a present vulnerability.

Questions readers ask