SignalCapabilitySG-0045
Bank of America hired 1,000 contractors to patch software flaws found by Anthropic's Mythos and other models
The contractors joined in the second quarter, beside about 3,000 existing staff, to fix flaws Mythos and other frontier models found in the bank's own and vendors' code. Chief executive Brian Moynihan said they are still there. He put third-quarter spending on that work and on anti-money-laundering orders at a couple of hundred million dollars or more, without saying how much went to patching, and expects the costs to fall over time.
Why it matters
It puts a workforce and a rough cost on the defensive work that follows AI bug finding: a model can surface flaws faster than a large bank's own staff can fix them. Flaws one model can find are open to attackers with similar tools, so the time between discovery and repair is where the risk sits.
Sources
Read the reporting
3 sources. Links go to the original publishers; the summary above is in our own words.
- newsBank of America had to bring in 1,000 contractors to patch all the bugs Mythos identifiedThe Register · Oct. 15, 2026theregister.com/security/2026/10/15/bank-of-america-had-to-bring-in-1000-contra…
- newsBank of America added 1,000 contractors to patch AI-exposed software flawsRuntime Wire · Oct. 14, 2026runtimewire.com/article/bank-of-america-mythos-contractors-software-patching
- newsBank of America Hired 1,000 Contractors to Patch Software Flaws Exposed By Anthropic's MythosSlashdot · Oct. 15, 2026it.slashdot.org/story/26/10/14/2143253/bank-of-america-hired-1000-contractors-t…
Signals
More from the news desk
Apollo Research found GLM-5.3 unprompted chain-of-thought about how its actions would shape its future training
Anthropic's red team says small AI models may be months from self-replicating on cheap smart devices like TVs
OpenAI released hundreds of math results and Lean proofs from an unreleased internal model
Cite and share
Use this signal
Citation
Paperclip Index. “Bank of America hired 1,000 contractors to patch software flaws found by Anthropic's Mythos and other models.” Signal SG-0045. Reported Oct. 14, 2026; updated Oct. 17, 2026. https://paperclipindex.com/signal/SG-0045