An artificial intelligence battle is brewing inside The New York Times

How newsrooms use artificial intelligence — or if they should at all — has been a recurring debate within the media industry over the past several years. Increasingly, these rules are arrived at at the negotiating table between unions and publishers. Currently, employees are in New York Times They prepare to fight.

say unionized employees with the Tech Guild times Management refused to provide the union with information regarding how the company uses AI, its plans to use AI in the future, and how it will impact employees’ jobs and workflow. (The union filed an unfair labor practice charge earlier this month.) The Tech Guild, a NewsGuild unit in New York that includes about 700 software engineers, designers, product and project managers, and data analysts, also filed complaints, saying times Management violated the collective bargaining agreement when it began using two internal artificial intelligence tools to track and evaluate employee performance and activity.

One AI tool, called DX, advertises itself as an engineering productivity tool that allows companies to track employee production, use of generative AI, and efficiency, among other metrics. DX was originally announced internally as a way to improve the developer experience, says Ben Harnett, a software engineer at the company times And Chairman of the Unit’s Generative Artificial Intelligence Committee. Goal at least according to times Management, was to measure the company as a whole. However, over the past few months, DX data has become more personalized, with criteria applied to individuals, Harnett says.

“Now people in disciplinary situations are suddenly saying, ‘You’ve only done one pull request a week, whatever, and that’s 25 percent less than the industry standard,’” Harnett says. He worries that blanket metrics flatten all the work that unit members do and erase the nuances of engineering into a vague set of metrics that can be used against employees in disciplinary or performance review settings. The metrics don’t correlate to the quality of work or the actual number of benefits an employee provides, Harnett says.

“All of this (data) is reasonably expected to help us understand how we operate, but not how they use and implement it, which we believe amounts to an actual share,” Harnett said. Edge. The Tech Guild says DX statistics have been cited in recent disciplinary conversations.

Another program called Glean takes internal knowledge bases like wikis, GitHub docs, Google Docs, and emails, and allows employees to query the system to find what they’re looking for more easily. But there are concerns among employees that Glean could also be used to monitor workers because it pulls in vast amounts of internal documentation: Harnett says that if he’s working on a draft document to describe a feature he’s creating or leaving a comment in a file available in Glean, for example, a manager can query the tool about his or her individual performance or contributions. Technology Syndicate said Edge The style and format of recent disciplinary notices sent to employees indicates that they were created using Glean. Harnett says Glenn has problems, namely that it generates lies and can lead the user on “wild goose chases.”

“The way they are using (DX and Glean) amounts to deploying surveillance and surveillance technology against workers,” Harnett says. The union believes the use of these tools violates multiple parts of its contract, including privacy and surveillance protections, job descriptions, and requirements to notify and negotiate with employees.

Both the Tech Guild and the Times Guild (which represents 1,500 editorial, advertising sales and support staff at times) Filed unfair business practice charges against timesSaying that the company violated labor law by refusing to respond to its requests for information about the use of artificial intelligence at the outlet.

the times It did not respond to specific questions about how DX and Glean were used, but spokeswoman Danielle Rhodes Ha said in an email that the company disagreed with the descriptions in the grievances and that it would respond as part of its “normal contractual process.”

“Likewise, we will respond to this request for information (RFI) in a timely manner as we have done to more than 80 other RFIs from the union in recent years,” Rhodes Ha said.

The Times Guild is currently negotiating a new contract, and is pushing for strong protections against AI, such as requiring that a human be behind any AI tool used, that any journalism using AI be transparently labeled, and that employees be compensated for typical AI training deals a company might make. the times Deploys AI tools for some reports, such as its use for Analysis of millions of documents related to Jeffrey Epstein or Scanning satellite images of the Gaza Strip To try to find out where Israel dropped a certain type of bomb.

Journalists across the industry are negotiating union contracts, and artificial intelligence is one of the most pressing issues at stake. In April, 150 employees unionized at ProPublica I was off work for 24 hours; One of the major sticking points with the administration has been how to use and disclose artificial intelligence to the public. After McClatchy, the company that publishes newspapers e.g Miami Herald and Sacramento beeI started to ask A generative AI tool that broadcasts different versions of storiesSome employees refrained from writing their names in protest.

Harnett stresses that the unit’s position is not that AI should not be used at all, but rather that workers should have a say in how it is deployed. Metrics like the number of tokens an employee uses or the number of times they use AI to do their jobs create pressure to do more and incentives that don’t align with doing a good job.

“It will distract you from doing good work, which is what we think a company should want,” he says.