We are the Nieman Journalism Lab, part of @niemanfdn at Harvard. We're trying to figure out the future of news.

Cambridge, Mass.
Joined August 2008
On Tuesday, two major German publisher associations called on the European Commission to prohibit Meta’s new terms of service from going into effect next week. They alleged the terms for Meta business accounts could violate the E.U.’s Digital Markets Act. niemanlab.org/2026/10/u-k-po…
7
3
12
2,120
The Wall Street Journal's custom AI tool tries to surface insights that traditional polling can miss. Called Doppler, it has already ingested 25 million TikTok videos and 500,000 podcasts ahead of the midterms. niemanlab.org/2026/10/the-wa…
1
2
1
7
1,535
Newsrooms may need to develop verification standards suited to information of illicit origin, transparency norms that treat uncertainty as something to disclose, and a shared language for the ethical judgments. buff.ly/blRFRWZ
1
1,156
We always try to ask the hard questions Thank you @HanaaTameez for a great interview!
NEW: I talked to the creators of @ASATNshow about comedy, explaining complicated stories, and why they traded in secure futures as business school grads for careers in making news videos. One of my fav interviews so far in the Creators of Record series! niemanlab.org/2026/10/two-cr…
1
2
4
914
NEW: After our reporting, Meta removed “train” from its upcoming commercial terms. But the terms still allow broad website scraping and could potentially supersede robots.txt. @decka227 and I followed up with the experts: niemanlab.org/2026/10/legal-…
11
3
25
3,284
On Friday, Meta removed the word “train” from a list of allowed use cases and limited those uses to “advertising and business products." The revised terms will still allow for broad scraping of websites, potentially superseding robots.txt files, according to legal experts. niemanlab.org/2026/10/legal-…
1
9
2
11
2,364
NEW: Virtually every news organization has a Meta business account. Meta’s new terms say it can scrape websites “associated with” those accounts—and use the content to “train” its products. Me + @decka227 on what this means for publishers: niemanlab.org/2026/10/metas-…
8
2
10
1,517