09-03-2025
CEOs "shoving" AI "into everything" — with mixed results
Companies across corporate America are experimenting with generative AI to see if it can make them better, smarter and more productive — with mixed success.
Why it matters: C-suite AI proponents have been pushing a "use it or get left behind" mentality, but it's often up to the rank and file to figure out how to actually implement AI in their day-to-day work.
What we're hearing: AI is helping workers offload time-consuming menial tasks, and it's handling some complex work better than humans.
Jason Rabinowitz, head of content creation at airline retailing firm ATPCO, told Axios that days of translating airline marketing content has been reduced to "about two hours" with AI's help in handling complex workflows and multiple spreadsheets.
Rabinowitz also described pitting AI-translated materials against human-translated versions in a "blind trial" — and finding that, so far, the AI-translated versions are "more readable and more accurate."
Reality check: Generative AI models suffer from "hallucinations" — techspeak for making stuff up.
AI's work needs to be checked, and that process is sometimes more time-consuming than not using AI at all.
And there remains the perennial concern among workers that generative AI will take people's jobs (with AI proponents countering that, like past disruptive technologies, it will create new, unforeseen jobs).
By the numbers: About 1 in 6 U.S. workers say they're using AI to do at least some of their work, per a recent Pew survey, while another 25% say AI could do at least part of their jobs.
52% of workers are worried about AI's impact, while 32% say it'll reduce job opportunities.
Yet 36% say they're optimistic about AI's potential.
The big picture: AI's value comes down to how it's used, says Alexia Cambon, senior director of research at Microsoft. (Microsoft is a major investor in ChatGPT maker OpenAI and runs a GenAI chatbot called Copilot.)
"There's a command-based approach, where you look at AI and you think, 'AI has to obey me — I'm going to give it a really simple prompt, and it has to do what I want it to,'" Cambon says.
"And then there's the conversation-based approach, which is ... 'I'm going to use it as a thought partner, and I'm going to use it to brainstorm' — and that requires a lot of critical thinking, and that is the preferable way to use AI in a work context."
The other side: Ed Zitron, CEO of PR agency EZPR and prominent AI skeptic, argues that many corporate leaders are pushing AI despite being too disconnected from their companies' day-to-day work to understand its actual use.
"What I think we're seeing is the biggest mask-off in corporate history, of bosses that do not know what they're talking about, that do not touch their businesses, shoving ChatGPT and other generative AI into everything because they don't know how anything works," Zitron says.
What's next: Generative AI proponents will tell you that the technology remains in its infancy and whatever comes next will be more capable.
The jury's out on whether that's true — hallucinations are an especially sticky problem.
But for anyone with an "email job," it couldn't hurt to at least start experimenting with AI to see what it can do for you — and what it can't.
Disclosure: Axios and OpenAI have a licensing and technology agreement that allows OpenAI to access part of Axios' story archives while helping fund the launch of Axios in four local cities and providing some AI tools. Axios has editorial independence.
