Banks are locking out AI Agents and shifting the losses to customers who let them in 🤖🏦; Visa wants Fireblocks for its neutrality, which a Visa takeover would end 🤷‍♂️💸

JPMorgan, Bank of America, Citi, Wells Fargo, Morgan Stanley and Schwab have restricted third-party AI agents that access customer accounts using shared passwords since mid-October, and Wells Fargo will begin treating an agent's actions as the customer's own from November 24, effectively shifting fraud liability under Regulation E. The blocking banks are also helping design Meta's official agent access path, while Visa's rumored $2.5B bid for Fireblocks would end the crypto custody firm's neutrality, a role the card network reportedly wants for stablecoin settlement.

Hey Everyone, Good morning & happy Monday Today, we’re diving into why JPMorgan, Bank of America, Citi, Wells Fargo & Morgan Stanley are locking AI agents out of customer accounts, and why customers who let them in may now carry the losses what the Axios scoop got wrong, why the blocks started a week before it, how Regulation E turns a shared password into a liability transfer, which blocking banks are helping design Meta’s official way in, & the one signal that would end the blockade + bonus deep dives into How AI Agents Reshaped Jobs, Hiring and Productivity at Anthropic https://linas.substack.com/p/anthropic-employment-report & How to Hire a Grok Bot: Your First AI Employee https://linas.substack.com/p/how-to-hire-a-grok-bot inside , and breaking down Visa’s rumored $2.5B bid for Fireblocks to see why the card giant wants the crypto custody firm for the one thing a Visa takeover would end: its neutrality what Fireblocks’ denial does and doesn’t rule out, why the price sits 69% below its 2022 valuation and who would take the loss, the stablecoin network gap Visa can’t close quickly on its own, & what to watch next + bonus deep dive into The Ultimate List of Stables Resources https://linas.substack.com/p/stablesresources inside . So let’s jump straight into the finnovative stuff 🌶️ Banks are locking out AI Agents and shifting the losses to customers who let them in 🤖🏦 The news 🗞️ The largest U.S. banks are now blocking AI agents that log into customer accounts with shared passwords , and customers who keep sharing them may be signing away their fraud protection. JPMorgan JPM 0.00%↑ https://substack.com/search/%24JPM , Bank of America BAC 0.00%↑ https://substack.com/search/%24BAC , Citi C 0.00%↑ https://substack.com/search/%24C , Wells Fargo WFC 0.00%↑ https://substack.com/search/%24WFC , Morgan Stanley MS 0.00%↑ https://substack.com/search/%24MS and Schwab SCHW 0.00%↑ https://substack.com/search/%24SCHW have all restricted outside agents since mid-October. Last week, Axios reported that banks had found agents inside customer accounts, some acting beyond user instructions. Its most alarming claims haven’t been verified yet , but regulators are already backing the blocks anyway, and from November 24, Wells Fargo will treat an agent’s actions as the customer’s own. Last month, we predicted big banks would block unidentified agents, build their own, and sell access later https://linas.substack.com/p/fintechpulse1132 . They’ve done the first two and are negotiating the third . So let’s take a closer look at why the blocks started before the scoop, how Regulation E turns a shared password into a liability transfer, which blocking banks are helping design Meta’s official way in https://linas.substack.com/p/fintechpulse1136 , and the one signal that would end the blockade.