Bank of America has hired 1,000 external contractors to help address software vulnerabilities exposed by Anthropic's Mythos AI model, according to CEO Brian Moynihan. The bank also spent over $200 million in the third quarter on software remediation and regulatory issues.
TL;DR
- Bank of America hired 1,000 contractors to patch software flaws exposed by Anthropic's Mythos AI model.
- The bank spent over $200 million in Q3 on software remediation and regulatory issues.
- AI is both a cost driver and a cost saver for the bank, with expected savings of $700 million by 2027 from internal AI projects.
What happened
Bank of America CEO Brian Moynihan revealed in a CNBC interview that the bank has hired 1,000 contractors to assist with patching software vulnerabilities. These flaws were exposed by advanced AI models, including Anthropic's Mythos, which was released to a select group of banks, technology firms, and government agencies in April.
The contractors are working alongside 3,000 internal staff to address both proprietary and vendor software flaws. Moynihan emphasized the real risk posed by these vulnerabilities, prompting the bank to take swift action.
In the third quarter, Bank of America spent a 'couple hundred million plus' on software remediation and regulatory issues related to anti-money laundering compliance, according to Moynihan.
Why it matters
The move highlights the dual-edged nature of AI. While AI can boost productivity and save costs, models like Mythos can also uncover and exploit decades-old software flaws, forcing firms into costly defensive measures.
For AI/ML developers and startup founders, this underscores the importance of robust software security and the potential risks of advanced AI models. Tech investors should note the significant costs and temporary nature of remediation efforts.
Despite the costs, Moynihan reassured investors that the spending increase is temporary. He expects remediation costs to decline over time as patches are applied. Additionally, the bank's internal AI projects are expected to generate substantial savings, rising from $300 million this year to $700 million by 2027.
Key facts
- Bank of America hired 1,000 contractors to assist with patching software vulnerabilities.
- The bank has 3,000 internal staff working on the same issue.
- Anthropic's Mythos AI model exposed the software flaws.
- The bank spent over $200 million in Q3 on software remediation and regulatory issues.
- Bank of America has about 130 internal AI projects.
- Expected savings from AI projects: $300 million this year, rising to $700 million by 2027.
- Bank of America's head count fell by about 3,400 employees in the past year.
- The bank cut roughly 5% of managers in the past year.
Context
Anthropic's Mythos AI model is part of a new generation of advanced AI models capable of uncovering and exploiting software vulnerabilities. This capability has significant implications for cybersecurity and software development practices.
The situation at Bank of America illustrates the broader impact of AI on enterprise operations. While AI can drive efficiency and cost savings, it also introduces new risks and challenges that organizations must address.
For the AI industry, this development underscores the need for robust security measures and the potential for AI to both create and mitigate risks. It also highlights the importance of ongoing investment in AI research and development to stay ahead of emerging threats.