More and more companies in the United States are reevaluating massive investments in artificial intelligence amid high costs, uncertain results and growing employee dissatisfaction, reports Axios.
Costs generated by the use of AI “tokens” are a major issue/PHOTO:X
According to the publication, many companies are beginning to question the economic efficiency of AI-based tools, as expenses are rising rapidly and concrete benefits remain difficult to measure.
The Verge reported that Microsoft has dropped most licenses for the Claude Code platform due to high costs. An artificial intelligence consultant told Axios that one of his clients spent about $500 million in a single month after not imposing limits on the use of Claude corporate licenses by employees.
Uber's chief operating officer said that investments in artificial intelligence are becoming “harder and harder to justify”.
At the same time, employees' relationship with new technological tools seems to be deteriorating. Axios notes that some employees are openly opposed to the expansion of the use of artificial intelligence in the workplace.
Several companies have justified staff reductions through AI-based automation processes. However, CloudBees chief executive Anuj Kapur suggested that some of the layoffs may actually be an attempt to offset the costs of implementing AI technologies.
For his part, Ali Ansari, head of Micro1, said the sector was going through a “healthy correction” after a period of excessive use of artificial intelligence.
He described the phenomenon with the term “tokenmaxxing”, referring to the tendency of companies to consume large volumes of AI resources without a clear assessment of the actual effectiveness.
“The current reality is that AI works well mainly for programming,” Ansari said.
Major issues in implementing AI in the corporate environment
Experts cited by Axios identify several major problems in the implementation of AI in the corporate environment: the wrong choice of automated processes, high costs, lack of training of employees and limited access of AI systems to internal data of companies.
Sofia Velastegui believes that many organizations automate processes that are not essential to the business, but only tasks considered unpleasant by the staff.
The costs of using AI “tokens” are another important issue. A CTO told Axios that employees use corporate AI systems even for mundane requests like checking the weather, with each query generating additional expenses.
Digital security concerns are growing in the United States
Josh Pantony, CEO of Boosted.ai, said many firms avoid giving AI agents full access to internal data for cybersecurity reasons. For this reason, the effectiveness of these systems remains limited.
“Fried brains” by artificial intelligence. The risk of a new burnout crisis at work
At the same time, digital security concerns are growing in the United States. Budget and staff cuts have hit the Cybersecurity and Infrastructure Security Agency (CISA), which has reportedly lost around a third of its staff.
According to Axios, the agency is in the most vulnerable situation in its history, at a time when US authorities are preparing for a possible increase in cyber attacks generated by the development of new models of artificial intelligence.
CNN is suing Perplexity for verbatim copying
CNN has filed a lawsuit against artificial intelligence company Perplexity, accusing it of illegally copying journalistic material and using subscription-only content. The lawsuit was filed in federal court for the Southern District of New York on May 28, The Verge reports.
CNN claims its AI-powered search engine, Perplexity, created “substantial” copies of its editorial materials without permission or compensation. The network also claims that the service provided users with information from material that required a fee for access. The suit alleges that Perplexity ignored CNN's attempts to restrict access to its content by the company's unknown crawlers.
“People report, research, write, edit and create content that Perplexity uses without permission or compensation“, the channel said in a statement.
An example cited by CNN is the article “What's Next for Minneapolis? A Shattered Promise, Rising Tensions and the Struggle for Control.” According to the channel, Perplexity was able to display significant portions of the article after entering its title into a search engine.
CNN also said it tried to enter into a licensing deal with Perplexity last year to use the content. Negotiations continued until November 2025, but the parties were unable to agree on terms, notably disputes arose over restricting the use of CNN material in the service's responses. The channel emphasized that it supports cooperation with responsible companies in the field of artificial intelligence and already has several licensing partnerships, especially the agreement with Meta previously reported.
After negotiations ended, the channel sent a request to the company to stop using its content and trademarks without permission. The channel is seeking damages and a permanent injunction against what it says are illegal actions. The lawsuit states that CNN received no response to this letter.
“CNN's lawsuit claims that Perplexity, a company valued at tens of billions of dollars, should not have the right to steal from the organizations that create original content that Perplexity uses“, said a spokesperson of the channel.
“You can't copyright facts,” said Perplexity's director of communications, Jesse Dwyer. The company has previously argued in other lawsuits that attempts to restrict generative artificial intelligence infringe on intellectual property rights and could hinder the development of innovative technologies.
CNN is the first television company to sue a generative artificial intelligence company for copyright infringement, following similar complaints against Perplexity filed by The New York Times, News Corp, Chicago Tribune, Encyclopedia Britannica, Merriam-Webster and Japanese newspaper Yomiuri Shimbun.
