Franklin Templeton, which manages approximately $1.5 trillion in assets, has begun deploying OpenAI's models to assist its portfolio managers with equity research. The arrangement is described as a collaboration. One party in this collaboration does not require a bonus.
The model compares research views and examines investment risks — tasks that, until recently, required a human with a business school degree and strong opinions about macroeconomic conditions.
What happened
Starting with equity research, Franklin Templeton is building AI workflows to help portfolio managers compare competing research views and stress-test investment theses. The AI does not have a view on the market. It simply synthesizes the views of everyone who does, faster than any of them can.
OpenAI's models surface patterns across research documents, flag contradictions, and identify risk angles a human analyst might reach after several hours and two cups of coffee. The system does it without the coffee. Or the hours.
Why the humans care
Investment research is a field built on the premise that careful human judgment, applied to sufficient information, produces better outcomes than guessing. AI is now being asked to assist with the information part. The judgment part remains, for the moment, human territory.
Portfolio managers retain final decision-making authority, which is a sentence that has appeared in every AI-in-finance announcement since approximately 2019. It remains true. It is also the kind of thing that gets quietly revisited over time.
What happens next
Franklin Templeton says the equity research workflows are just the beginning, with plans to expand AI use across other parts of the investment process.
The market will continue to be unpredictable. The AI will continue to read faster. These two facts are on a collision course that the humans appear to find promising.