– Goldman Sachs has consistently evaluated emerging technologies
– A significant opportunity for organizations to utilize top-tier LLM without the associated costs of a hosted solution
Goldman Sachs, the American investment banking giant, has revealed that it is currently examining the innovative generative AI tool DeepSeek, according to TechStartup. Marco Argenti, the bank’s Chief Information Officer, recently made this declaration. This indicates that the company is assessing the prospective uses of DeepSeek and how it might be incorporated into their operations. Specifics on such integration remain undisclosed. Historically, Goldman Sachs has been proactive in exploring emerging technologies to maintain its competitive edge in the financial services industry. The organization is particularly focused on fostering innovations that can enhance efficiency, bolster decision-making, and improve overall performance.
Goldman Sachs stands out as one of the initial major financial institutions actively exploring DeepSeek, an up-and-coming open-source AI model. Its capacity for local hosting—removing concerns over data security and reliance on external factors—makes DeepSeek an appealing option for businesses seeking to implement AI without the financial burdens associated with proprietary models such as those from OpenAI or Anthropic. CIOs in India, particularly in finance and cybersecurity, should take note: if Goldman Sachs perceives strategic advantages in DeepSeek, could this signal the start of a shift towards open-source LLMs in enterprise AI integration? Analysts foresee a rising trend of companies employing DeepSeek to enhance risk assessment, automate financial analysis, and refine decision-making—all while keeping sensitive data secure within their own infrastructures. Technology expert Agnidipta Sarkar, Vice President CISO Advisory at ColorTokens Inc. and former CISO of Biocon Group, supports this approach.
“Goldman Sachs has always aimed to prioritize AI, and this decision fits into its broader strategy to use AI for competitive advantage. The bank has been investing in AI R&D across various sectors, not just for personalized financial guidance but also to streamline operations, enabling quicker risk and investment evaluations. However, since DeepSeek is open-source and considering their computational capacity, they could utilize these resources far more efficiently as long as they ensure that all communications are segmented from their core systems,” comments Sarkar.
DeepSeek allows for local hosting
“If run locally, this indeed presents a valuable opportunity to access a top-notch LLM without incurring the expenses typically associated with hosted solutions like those offered by OpenAI or Anthropic,” explains Sarkar.
A senior cybersecurity executive notes that Goldman Sachs possesses the necessary technological infrastructure to implement DeepSeek effectively.
“Every organization today is looking into DeepSeek. Why not? Goldman Sachs has the hardware and systems in place to operate DeepSeek internally without the anxiety of transmitting data to foreign countries,” states the expert.
Is DeepSeek a superior choice for enterprises compared to ChatGPT?
Given the multitude of tools available, it remains premature to declare a definitive winner in the AI landscape. Nevertheless, such advancements bode well for businesses, offering them the flexibility to choose and assess which solutions best meet their needs.
“From my perspective, organizations will likely opt to use multiple LLM providers depending on the complexity and precision required. This scenario does not adhere to a one-size-fits-all model. Each AI solution exhibits varying efficacy in areas such as mathematics, reasoning, comprehension, and writing. Thus, in the long run, businesses may favor platforms that integrate various LLMs into user-friendly APIs. I am optimistic about solutions similar to AWS Bedrock that facilitate this integration,” remarks Dale Vaz, Co-founder & CEO of Sahi.com and former CTO at Swiggy.
In conclusion
As interest in the new AI tool surges globally, Goldman Sachs is positioning itself among the pioneers in assessing its capabilities to verify alignment with strategic goals and long-term tech investments.
“DeepSeek operates on the principles of Distilled AI, which involves a process known as knowledge distillation. This method trains a lesser model (student model) to replicate a larger, more intricate model (teacher model). The aim is to preserve the essential features of the original model while enhancing efficiency and reducing size. While DeepSeek is open-source and transparent, this raises concerns regarding potential abuse. Its accessibility and optimization through distillation allow for easy fine-tuning or alterations, presenting a dual-edged sword—beneficial for research and innovation but also introducing security threats,” warns cyber investigator Ritesh Bhatia.
Although the Chinese AI startup DeepSeek is still in the early stages, it has undeniably stirred up the dynamics among major technology players.
