Colgate AI head is Princeton phd in Chemistry

Colgate’s AI Evangelist Teaches It to Embrace Disruptive Future 

BY BRADLEY OLSON

 

 

Iraklis Pappas couldn’t let it go.

On a conference call last April, an ad agency partner began presenting unpolished AI-generated images to make a point: Artificial intelligence wasn’t ready to take center stage in advertising.

Pappas, the global head of AI forColgate-Palmolive , a 220-year-old consumer-products company, quickly interjected, pointing out that the agency was using an older tool. He took over screensharing, showing Colgate executives how a newer Chat-GPT image model was far more capable.

AI is misunderstood, said Pappas, who has worked at Colgate for 15 years and goes by “Kli.” “There’s the straw man of, ‘Well, it failed at this one thing, therefore it’s all stupid.’” Such interactions—in which Pappas bluntly challenges what he deems anti-AI sophistry— occur regularly as Pappas acts as a kind of AI evangelist at Colgate, whose brands include its eponymous toothpaste and soap as well as Speed Stick deodorant, Ajaxcleaners and Hill’s Pet Nutrition. Often, he said, he walks the halls in New York and New Jersey offices in search of AI tinkerers whom he can turn into companywide megaphones helping to spread the good word.

Pappas, 38 years old, and his team have been at the heart of a surge in AI usage among Colgate’s white-collar workforce. Colgate says 51% of its salaried and clerical workers used advanced artificialintelligence tools weekly by the end of last year, up from about one-third shortly after OpenAI released ChatGPT in late 2022. They got there through a combination of Pappas’s debate-style instruction, myth-busting and formal workshops and hackathons.

Pappas is a foot soldier in an increasingly high-stakes battle taking place across American corporations. Many

 

 

white-collar workers outside of software engineering have been slow to embrace fastchanging AI tools.

OpenAI co-founder Sam Altman, Microsoft Chief Executive Satya Nadella and others have spoken about a yawning gap between AI-model capabilities and individual usage. While frontier startups and tech giants are planning to spend trillions of dollars in coming years to build data centers that can meet the computing needs of AI systems, that spending is largely based on assumptions that millions more users—especially corporations—will pay for AI tools and services.

Coders have expressed awe at the extent to which powerful AI models can replicate their expertise. But rank-andfile employees across many other industries have been slower converts due to fears of being replaced, a lack of time for training and frustration typical of any population learning new technology.

The potential for disruption across industries has rattled markets in recent weeks, but economists and others are divided in their forecasts about AI’s effect on productivity and profits.

Pappas dismisses as “media hype” claims that AI adoption isn’t a profit driver. Instead, he steers Colgate employees to an academic study currently under peer review showing AI’s positive impact on the bottom line at an online-shopping platform, and he points to internal Colgate survey data showing positive nonmonetary effects.

“Every single investment has a business case” at Colgate, Pappas said. Before spending money on software, he is required to estimate the returns.

Pappas began at Colgate as an intern while an undergraduate at Rutgers University in New Jersey and left for four years to earn a Ph.D. in chemistry from Princeton University. Before becoming the company’s head of AI in 2023, he was a data scientist and worked on predictive analytics.

Pappas formed a partnership with one of Colgate’s top executives, lawyer Nadine Flynn, to smooth the path for rolling out AI software to employees.

They added “responsible AI principles” to Colgate’s code of conduct, making clear the goal wasn’t to replace workers with AI but to “augment them,” Pappas said. The revision triggered mandatory employee training.

One challenge for executives pushing AI on their teams is that employees fret over being sidelined by the technology.

“Across industries, there either is or should be some level of fear for companies and for individual employees,” said Gretchen Greene, the chief of artificial intelligence at law firm Ropes & Gray. But “if we don’t learn to use AI well, those who do will get ahead.”

AI has helped Colgate find savings in a number of unexpected places. One was in Greece.

A plant in Athens that produces homecare products was plagued by downtime, with production idled for days, when manufacturing e q u i p m e n t broke because manuals were in German, French and English. The plant manager uploaded the manuals and built an AI assistant that could translate and answer questions in Greek, allowing employees to get machines back up and running faster.

The company expanded the tool to handle a variety of manuals and documents in many languages and has rolled it out to 43 manufacturingand production sites globally.

Even with success stories like that, Pappas and his team have faced frustrations in their efforts to drive more AI usage. For the first few years, Pappas says he just saw the same pool of employees—about 30% of white-collar workers—engage, but more hesitation from potential new converts.

Colgate doesn’t count simple tasks, such as having AI help create PowerPoint presentations, in its statistical tracking of employee AI use. Instead, Pappas is pushing for activity like deep research and coding in Google’s Gemini, ChatGPT Assistants or NotebookLM, a Google AI product that can analyze specific documents or other resources.

To reach additional workers, Colgate created an “AI ambassador” program. Twenty-four people representing every major business unit work closely with Pappas. Each of them in turn manages another 10 ambassadors within their departments.

Now, more than half of white-collar workers log “advanced” AI use weekly.