Goldman Sachs ‘Hyperaware’ of AI Risks; Collaborates with Anthropic on Mythos Model
Goldman Sachs is adopting a cautious stance regarding a new artificial intelligence model developed by Anthropic, known as “Mythos.” The financial institution has raised alarms about the model’s advanced capabilities, which could pose significant cybersecurity risks, even as it explores the model’s long-term potential.
Mythos has generated considerable concern within the financial sector due to its ability to autonomously identify and exploit software vulnerabilities. This capability has the potential to fundamentally alter both cyber defense strategies and the landscape of cybercrime.
Understanding AI-Driven Cyber Risks
During a recent earnings call, Goldman Sachs CEO David Solomon emphasized the bank’s vigilance regarding the risks associated with emerging AI systems, particularly large language models (LLMs) and the disruptive Mythos model from Anthropic. Solomon stated, “We’re hyperaware,” highlighting the cybersecurity implications of next-generation AI tools.
Goldman Sachs is actively collaborating with Anthropic and various cybersecurity partners to gain a deeper understanding of how such models could impact financial systems and cyber defenses. Solomon noted that cybersecurity has long been a core focus for the bank, which has invested substantial resources into this area.
“With the help of the US government and the model publishers, we are very focused on supplementing our cyber and infrastructure resilience,” he added. This reflects a growing trend among major financial institutions, which are increasingly viewing advanced AI not merely as a productivity tool but as a potential disruptor to security.
The Concerns Surrounding Mythos
Mythos distinguishes itself from earlier AI systems by its design to autonomously discover and exploit vulnerabilities in software environments. Anthropic has acknowledged that the model can “find and exploit sophisticated vulnerabilities,” sometimes outperforming human experts in this domain.
This capability has raised alarms within the cybersecurity community, with experts warning that such tools could lower the barrier for cyberattacks. Even individuals lacking deep technical expertise could potentially leverage AI to identify weaknesses in operating systems, applications, or enterprise infrastructure.
Anthropic has taken a cautious approach by restricting access to Mythos and opting not to release it publicly, citing concerns about potential misuse. Instead, the model is being shared as a preview with 11 organizations under a controlled initiative called “Project Glasswing.” This initiative includes major players like JPMorgan, Apple, Google, Microsoft, Nvidia, and Goldman Sachs, aiming to strengthen defenses before broader deployment.
Financial Sector on High Alert
The concerns regarding AI-driven cybersecurity risks extend beyond Goldman Sachs. High-level discussions among U.S. financial leaders, including regulators and central banking officials, have reportedly taken place to assess the risks posed by such AI systems. Banks are particularly vulnerable due to their complex mix of modern and legacy systems, which may provide fertile ground for AI-driven vulnerability discovery and exploitation.
Industry leaders recognize a dual-edged reality: while attackers may initially benefit from such advanced tools, defenders could eventually adopt similar technologies to identify and patch weaknesses more rapidly.
Balancing Risk and Opportunity
Despite the warnings, Solomon maintains a measured perspective on the future of AI in business. He noted that the technology has the potential to significantly enhance efficiency and transform operations across various industries. “Whenever you have acceleration of your technology, there are going to be bumps, and there are going to be risk issues,” he stated. “But the power of the technology, the ability to use it in an enterprise, to remake processes, to create efficiency, and also create more capacity to invest in growth — I can’t find a CEO that’s not talking about that.”
This tension between innovation and risk is central to the ongoing debate surrounding advanced AI systems like Mythos.
A Turning Point for Cybersecurity
The emergence of models capable of autonomously identifying and exploiting vulnerabilities signifies a potential inflection point for cybersecurity. Experts suggest that the rapid evolution of AI could accelerate both offensive and defensive capabilities, creating a race between attackers and defenders. In the short term, however, the concern remains that powerful tools may be easier to weaponize than to secure.
For financial institutions like Goldman Sachs, the strategy appears to be one of early engagement, understanding the risks, and preparing defenses before such technologies become widely accessible.
Source: thecyberexpress.com
Keep reading for the latest cybersecurity developments, threat intelligence and breaking updates from across the Middle East.
