@alixdunn.com @doriantaylor.com @davekarpf.bsky.social
18.02.2026 22:28 — 👍 1 🔁 0 💬 0 📌 0@subbuvincent.bsky.social
I direct Journalism & Media Ethics at Santa Clara University’s Markkula Center for Applied Ethics. Focus: sourcing standards; role of algorithms/AI in news distribution; democratic culture. https://www.linkedin.com/in/subbuvincent/
@alixdunn.com @doriantaylor.com @davekarpf.bsky.social
18.02.2026 22:28 — 👍 1 🔁 0 💬 0 📌 0
The confidence makes people dismantle safeguards, so watch out.
This Y Combinator thread went on a blast over the weekend (T/H my friend Rohit Khare for flagging it to me.)
news.ycombinator.com/item?id=4702...
And the actual retraction
arstechnica.com/staff/2026/0...
With secondary work, the smooth and confident responses from AI bots is a removal of friction that slips the brain. Worldwide, tons of writers might be quoting people from secondary text and might simply be falling for how the bots respond so confidently.
18.02.2026 22:28 — 👍 0 🔁 0 💬 1 📌 0
Primary sourcing with people is usually direct line of communication. You may also feel a greater responsibility to your source, and may send their quotes back to them to double-check. More human investment overall.
The risk is higher for secondary sourcing because of how AI works.
For primary sources, precisely because you as a reporter found something your source said was worth quoting, your brain (memory) is also likely to give you at least faint warning that the AI tool is producing a “quote” that you did not *actually hear* when you talked to the person.
18.02.2026 22:28 — 👍 0 🔁 0 💬 1 📌 0But there is more to this: the risks are not the same for primary and secondary sourcing. When journalists talk to a source directly, it is primary sourcing. Where you pull from a source’s speech already posted elsewhere, in this case a blog, it is secondary sourcing.
18.02.2026 22:28 — 👍 0 🔁 0 💬 1 📌 0It is a cardinal lesson. The basic safeguard Edwards said he ought to have followed - simply comparing the generated text with the lines he wanted to pull from the blog post verbatim - would have helped catch the quotes error. The key thing is you have to apply that safeguard all the time.
18.02.2026 22:28 — 👍 0 🔁 0 💬 1 📌 0
Last weekend, Ars Technica retracted an AI-involved-news story by reporter Benj Edwards because it had fabricated quotes. Edwards has already explained how it happened, taken responsibility, and said that he had asked his boss to pull the piece.
bsky.app/profile/benj...
Yep it is. And this type of inquiry is doable.
I don’t know of published datasets that combine source attributions with analysis of language for promotion/PR/framing for tech companies, but that’s a valid hypothesis already.. will check
So many people are asking how they can support me..
The best way .. is to support @resistanceschool.bsky.social!
I’ll be teaching Race, Media and International Affairs 101 and 102 next month.
Sign up for lectures by me and guest speakers!
www.resistancesummerschool.com/fall-2025-re...
Statement from @postguild.bsky.social on the firing of Karen Attiah:
15.09.2025 19:01 — 👍 6981 🔁 1675 💬 126 📌 52"The executive branch cannot be both the arbiter of legality of media content, and the issuer of takedown notices" @subbuvincent.bsky.social says as @reuters.com documents scope of Indian government crackdown - including on inconvenient news, satire, cartoons, etc. www.reuters.com/business/med...
06.08.2025 08:35 — 👍 1 🔁 1 💬 0 📌 0All star group on how journalism is failing to respond to authoritarian elites, fake news, and AI--and how it can pivot to meet the moment. @subbuvincent.bsky.social @jayrosen.bsky.social @joelsimonsays.bsky.social and others. www.cjr.org/feature/thir...
21.08.2025 17:55 — 👍 2 🔁 2 💬 0 📌 0
Great piece by @subbuvincent.bsky.social on avoiding both-sidesism and ignoring fake controversies!
"Once you take somebody else’s framing, and you make it part of your controversy, you’ve legitimized the controversy... A better approach is: Determine whether something is a controversy or not."
Subbu's piece builds on the framework he developed in this June 2024 piece on pro-democracy journalism. kettering.org/to-counter-a...
22.07.2025 16:56 — 👍 0 🔁 1 💬 0 📌 0New on our blog: How can news media resist authoritarian efforts to control the narrative? Journalism scholar @subbuvincent.bsky.social shows a new model is starting to emerge in recent coverage of the Trump admin.
22.07.2025 16:53 — 👍 6 🔁 4 💬 1 📌 0News headlines are the first and most important interface between readers and journalists, but few news organizations address the mechanics of headline writing in their ethics codes. Read our latest with @subbuvincent.bsky.social on his new guidance for newsrooms balancing ethics and SEO:
01.07.2025 19:04 — 👍 2 🔁 1 💬 0 📌 0Yes Andy, and thank you. I forgot to give Paul and Burt the link to insert into the schedule. I’ll do that today. DM me your email and I’ll send you the deck anyway.
08.05.2025 13:39 — 👍 1 🔁 0 💬 0 📌 0I’m looking to leading this session and meeting fellow Hacks/Hackers summiteers at Baltimore soon!
21.04.2025 17:04 — 👍 2 🔁 0 💬 1 📌 0JSK alum Paulette Brown-Hinds and her team at Mapping Black California build data visualization tools that expose systemic inequalities and hold institutions accountable. How they give communities the necessary information to push for real change...
22.02.2025 16:08 — 👍 1 🔁 1 💬 0 📌 0
I've resigned from the jury for the duPont-Columbia Journalism Awards in response to Columbia's capitulation to Trump.
www.poynter.org/commentary/2...
Great Q. There is no openly and transparently justified value-add, it just isn't in the premise. It seems to come from a misplaced assumption about the worth of journalistic power itself. That reflexive neutrality to describe power is a higher norm than using democratic culture as the yardstick.
19.03.2025 05:28 — 👍 2 🔁 0 💬 0 📌 0
And that unfinished sentence is still there this afternoon, March 17th, 2025.
Perhaps they find the sheer acts of documenting executive orders and actions in plain written word too bothersome?
And I wrote about this very thing here. And I explained why journalistic practice owes democracy every day.
kettering.org/to-counter-a...
t/h @carlamurphy.bsky.social
Hi Surya, would you DM me you email? I saw the CITP article on your work (Feb 14) and liked the rare point you made about giving people agency. Would like to check in with you more, and include your views in stuff I might write. I am an engineer like you who got into journalism and much later ethics
21.02.2025 04:13 — 👍 2 🔁 0 💬 0 📌 0
UT Dallas tried to squelch the students who ran the campus newspaper. They decided to do their own thing and launch an independent counter-news outlet. American muckraking at work?
www.texastribune.org/2025/02/07/u...
This feels like eons ago now, and so much water down the rivers since then!
10.02.2025 21:31 — 👍 4 🔁 0 💬 1 📌 0Thank you for the reply Eric. This is a chronic editorial ethics issue that plagues journalism.
07.02.2025 18:25 — 👍 2 🔁 0 💬 0 📌 0
Eric, did they allow you to sign off or veto the final edited version of your piece? I.e. did you have the option to say NO to running such a piece with your name on it, given the editing?
(I'm considering making this a use case for my work on editorial policy and I wanted to ask you this first!)
Neutrality has a role in journalism when the facts are legitimately in dispute. Fact-finding and diverse sourcing need to determine the truths on the ground.
13.01.2025 15:29 — 👍 5 🔁 1 💬 1 📌 0