Plus: Misinformation swirls around White House Correspondents’ dinner shooting; Adidas makes hay of marathon record smash.
In the early days of tech adoption, many companies will heavily subsidize their products to encourage sampling. During the early days of Uber, for instance, the company gave rides at a loss just to get people to try the technology and see how much more convenient it was than taxis. Then, over time, those subsidies are removed — and sticker shock can follow.
We may now be entering that era with AI.
Marketplace reports that the cost companies are spending on AI has nearly doubled in the last year and is now rivaling the cost of just paying a human.
Part of the issue is that companies are relying more on AI, certainly. But as AI grows smarter and more complex, it also takes more resources to run. Using a simple chatbot takes more tokens — read, money — than a semi-autonomous agentic AI. But that AI agent has far more productivity upside, so many think it’s worth the cost.
But some AI companies, including Anthropic and Microsoft, are changing their pricing structures to reduce subsidies, which could mean you’re paying more for that productivity.
It may no longer be nearly as cost-effective to replace a software engineer with a bot: Goldman Sachs says the cost of running a high-tech agentic AI could soon rival the salary of that engineer — though admittedly without the need for benefits, PTO or lunchbreaks.
Why it matters: The early adoption phase of AI is quickly coming to an end. Companies that invested big now want to see big returns. That money has to come from somewhere. But will AI soon seem like such a good deal if it costs as much as a human? It will likely depend on the exact skillset required and how many humans are needed to oversee the technology.
At the same time, many humans are beginning to backlash against AI, upset over its capacity to replace jobs or turn human creativity into “slop.”
For companies messaging their AI use, it’s important to think strategically. The time when you could get lots of attention for just slapping an AI label on a product is long gone. Now people are more interested in knowing how humans were involved. That’s becoming the true selling point. And as the cost of using advanced AI draws near to the cost of hiring a skilled human, we may begin to see an even greater change in the overall messaging environment around AI.
Don’t be left behind in the pivot.
Editor’s Top Reads:
- Over the weekend, a gunman crashed the White House Correspondents’ dinner, where President Donald Trump was in attendance. While one law enforcement office was shot, he was saved by a bulletproof vest. The gunman was taken into custody. As soon as word of the shooting broke, rumors and conspiracy theories began to spread heavily online. Many were skeptical of whether the shooting was staged. Others debated whether the shooter had been killed on scene. It’s an unfortunately familiar pattern for any major event occurring today. As Cliff Lampe, professor and associate dean for academic affairs at the School of Information at University of Michigan, told the New York Times: “People are reshaping reality based on what they want to be true or not. They’re not looking for good information, they’re looking for confirmatory information, and will often go very deep down a rabbit hole of side-by-side pictures, microshots of the president’s face, et cetera.” This is true of a shooting incident and true about rumors of your brand. Be aware that in the wake of any incident involving your brand, the internet will quickly take over and lose control. Social listening, prebunking information when possible and quick responses are the only way to hope to combat misinformation. Have a plan now.
- During the London Marathon this weekend, three records were set. For the first and second time, male runners ran the 26 miles in less than two hours. The winner of the women’s event also broke her own record. The common denominator, besides superhuman strength that makes me feel exhausted just thinking about it? Their shoes. The Adidas’ Adizero Adios Pro Evo 3 is gaining a significant amount of press in the wake of the incredible accomplishments. The shoes are described as lighter than a bar of soap and new world record holder Sabastian Sawe called them the best shoes he’d ever worn. Adidas was ready to take advantage of the situation, with their VP of running ready to discuss the high-tech that went into making the shoes. Their homepage is currently hosting a takeover of Sawe and the Adizero shoes, bearing the tagline “History in Time. Humanity just got faster. Powered by ADIZERO.” The engineering team had to make the shoes, Sawe had to run his heart out and the comms and marketing team were equally on top of their game to capitalize on the outpouring of interest. Now, they’re earning headlines like “Sawe’s marathon record boosts Adidas in ‘supershoe’ race with Nike” — the stuff PR dreams are made of.
- OpenAI CEO Sam Altman has made a public apology to the Canadian community of Tumbler Ridge after a young woman killed eight people and herself. The shooter was banned from ChatGPT in June 2025 for activity that gave “an indication of potential real-world violence.” But OpenAI did not report her to authorities, and she was able to create a second account, which was not discovered until after her death. Most of the letter, published in a community news site, is standard thoughts and prayers fare, as well as explaining why he had delayed several months in releasing the apology. But at the end, he makes a concrete promise: “I reaffirm the commitment I made to the Mayor (of Ridgeline) and the Premier (of British Columbia) to find ways to prevent tragedies like this in the future. Going forward, our focus will continue to be on working with all levels of government to help ensure something like this never happens again.” It sounds strong, but it lacks concrete details on what that work will entail. Will anyone whose account is closed for indications of violence be reported to the police? How will the company balance privacy requirements with a need for safety? It’s ultimately a statement that raises far more questions than it answers — and which will require extensive follow-up supported by action.
Allison Carter is editorial director of PR Daily and Ragan.com. Follow her on LinkedIn.
The post The Scoop: The cheap all-you-can-eat AI buffet may be coming to an end appeared first on PR Daily.
