Tong Li
@tongli.bsky.social
Digital Media & Brand Strategy Expert | Former Journalist & Skilled Storyteller | Data & Business Analytics Enthusiast
Ads in ChatGPT aren’t inherently the problem. Blurring ads with AI judgment would be. Clear labels, privacy, sensitive-topic safeguards + answers that stay independent of advertisers should be the baseline. In conversational AI, trust is the product. www.bleepingcomputer.com/news/artific...
OpenAI will show visual ads in ChatGPT while you generate images
OpenAI is expanding ads in ChatGPT, and one of the first new formats will show visual ads while you're generating images.
bleepingcomputer.com
$10M for serious AI journalism is welcome. AI companies have enormous money + political influence; the reporters scrutinizing them need resources too. But funding should come with disclosure and strong editorial firewalls. www.semafor.com/article/10/0...
Exclusive: AI nonprofit will spend $10 million on journalism
A Silicon Valley nonprofit close to the AI safety movement is funding coverage of artificial intelligence.
semafor.com
Surveillance of journalists isn’t just a privacy issue. If sources assume every call or message can expose them, some will simply stop talking. Press freedom requires more than the right to publish. It requires space to report without being watched. www.ifj.org//media-centr...
New global IFJ report urges stronger regulation of surveillance / IFJ
The International Federation (IFJ), the largest journalists’ organisation across the world, has launched its latest report, Hacked Silence: Journalists and surveillance- addressing legal and regulator...
ifj.org
AI licensing needs more than checks. It needs measurement. If publishers can’t verify when their journalism was retrieved, cited or used to ground an answer, they can’t negotiate its value. A shared standard is progress—but Big Tech can’t be the only one keeping score. digiday.com/media/spur-p...
SPUR publishes AI content tracking standard, pitches OpenAI and Google to join advisory board
Publishers are pushing for AI transparency with a new SPUR standard to track how AI tools use content, and inviting frontier AI labs to help.
digiday.com
Creators using AI isn’t the same as AI replacing creators. Use it for analytics, editing and busywork? Great. Let the algorithm decide the hook, tone and structure of everything? That’s how creativity becomes sameness. AI should expand human expression—not standardize it. digiday.com/media/the-cr...
The creator economy’s stance on AI is shifting — again
Creators are integrating AI tools into their workflows, but at what point does it overtake the creative process and result in sameness?
digiday.com
Reddit’s bot problem is real. Killing RSS and public API access is still the wrong kind of fix. It blocks scrapers—but also researchers, moderators and ordinary people who use open tools. techcrunch.com/2026/09/30/r...
Reddit is killing RSS feeds and ending public API access because of AI bots | TechCrunch
Reddit is ending support for RSS feeds, as the company continues tightening access to its trove of user-generated content.
techcrunch.com
The Inquirer’s Scrape project gets AI in local news right: use it to surface leads, not write the story. A tool that cuts 12 hours of source-scanning while editors still decide what matters is augmentation, not replacement. www.lenfestinstitute.org/solutions-re...
How The Philadelphia Inquirer built Scrape, an AI tool to surface hyperlocal news leads
The Philadelphia Inquirer's AI tool, Scrape, combs through hyperlocal news leads, enabling editors to report on a broader geographic area.
lenfestinstitute.org
Good journalism isn’t automatically a sustainable business. If an independent newsroom survives because one founder does six jobs, works unsustainable hours or relies on a spouse’s income, that’s not resilience. www.poynter.org/business-wor...
What does it take to make an independent local newsroom work? More than good journalism. - Poynter
A new study of 112 independent publishers shows just how much local news entrepreneurs have to juggle
poynter.org
California’s new local-news law ties tax credits to newsroom jobs: hire more reporters, get more support; cut staff, get less. But hedge funds can qualify too. The real test is whether public money produces more original reporting where communities need it most. www.niemanlab.org/2026/10/cali...
California’s big local news bill becomes law
At the end of August, the state of California passed a bill to give news organizations tax credits based on the number of journalists they employ. It's the largest local news bill in the United States...
niemanlab.org
Columbia Journalism School pausing its M.A. while launching its first online degree feels like a bigger question about who journalism education is built for. www.niemanlab.org/2026/10/colu... #Journalism #JournalismEducation
Columbia Journalism School will pause its M.A. program (and launch an online degree)
The Columbia Journalism School will pause admissions for its Master of Arts program, dean Jelani Cobb announced on Thursday. The country's most prestigious journalism school also announced a new onlin...
niemanlab.org
Google reportedly is paying ~100 publishers when their journalism contributes to AI Overviews, AI Mode and Gemini. Good principle: If reporting creates value for an AI answer, the newsroom that financed that reporting should share in the value. www.theverge.com/tech/1002665... #Journalism #AI
Google reportedly tests paying publishers for AI search results
Around 100 publishers have joined Google’s pilot program, The Information reports.
theverge.com
The most important audience shift in journalism may be moving from: “How do we get more clicks?” to: “How do we become something people choose to return to?” www.niemanlab.org/2026/09/how-... #Journalism #AudienceEngagement
How publishers are rethinking audience engagement: Five takeaways from the Future of Media Technology Conference
"It would be a mistake to try to replicate the creator economy."
niemanlab.org
Big AI copyright ruling — but an important caveat: The appeals court ruled against Ross Intelligence, which used Westlaw headnotes to build a competing legal search product. The court stressed that Ross was't a generative AI system and did not create new expression. www.niemanlab.org/2026/09/fede...
Federal appeals court upholds Thomson Reuters’ landmark AI copyright win
On Tuesday, a U.S. federal appeals court ruled in favor of Thomson Reuters, upholding the company's landmark win in a copyright infringement case against the now-defunct legal AI startup Ross Intellig...
niemanlab.org
I like The Guardian’s approach to “journalists as creators.” It’s creating 40 paid roles with structured training in video, audio, photography, social, text + audience engagement — with at least 20% of work time dedicated to training. pressgazette.co.uk/news/guardia... #Journalism #CreatorEconomy
Guardian creates 40 jobs to train journalists as content creators
The Guardian is creating 40 new jobs which will feature training to help newer journalists gain content creator skills.
pressgazette.co.uk
NPR becoming “video-first” sounds almost contradictory. But maybe that’s the wrong way to think about it. NPR’s mission was never “make radio.” It was public-service journalism. www.niemanlab.org/2026/09/npr-... #PublicMedia #Journalism
NPR, the quintessential audio news brand, adapts to a video-first future
NPR’s new chief content officer Nadine Zylstra on how the public media company acquired Machine Gods — which she thinks can be “the Car Talk of AI.”
niemanlab.org
“AI shame” sounds like an individual problem. Nearly 70% of surveyed communication scholars already use generative AI somewhere in their work — but many still worry that admitting it will make them seem less credible. www.niemanlab.org/2026/09/a-la...
A lack of professional AI guidelines leads to feelings of “AI shame,” a new study finds
Are you a journalist who has used AI for work in some capacity but been nervous to tell anyone about it? Worried about how your bosses might see you, or that your work might be discredited? Well, aca...
niemanlab.org
This is much closer to the AI investment model I want for local journalism. OpenAI is adding $10M in cash + credits to Lenfest’s fellowship program — but instead of funding an “AI writes the news” product, it embeds full-time technical staff inside local newsrooms. www.niemanlab.org/2026/09/open...
OpenAI doubles support for Lenfest’s AI and local news fellowships
With an additional $10 million in program funding, the Lenfest Institute will consider new applicants in local TV news and statewide public service journalism.
niemanlab.org
The NYT says it wants three things from AI licensing deals: Fair value. A sustainable partnership. Control over how its journalism is used. Good principles. But The Times can walk away from a bad deal. A small local newsroom often can’t. digiday.com/media/what-t... #Journalism #AI
What The New York Times looks for in AI licensing partnerships
The New York Times is seeing a more publisher-friendly AI licensing ecosystem, but the marketplaces still have a long way to go.
digiday.com
A disaster newsroom doesn’t just need reporters. It needs water, power, food, working phones, a low-bandwidth website — and people who haven’t been awake for 36 hours. That’s why I like Emergency Mode for News. www.niemanlab.org/2026/09/when... #LocalNews #Journalism
When disaster strikes, newsrooms scramble. A new website aims to help tame the chaos.
Emergency Mode for News has checklists and other tools to help journalists respond to disasters before, during, and after they happen.
niemanlab.org
The most important revelation here isn’t an inflammatory internal quote. It’s the “doom loop.” Microsoft employees reportedly worried that AI could depend on human-created content while simultaneously destroying the economic foundation of the people who produce it. www.niemanlab.org/2026/09/an-a...
“An astonishing theft of unprecedented proportions”: Court records show what Microsoft and OpenAI actually thought about AI training
Would you call large language models training on an internet's worth of content "an astonishing theft of unprecedented proportions"? Or perhaps you prefer the "largest theft of labor in human history"...
niemanlab.org
The DOJ siding with OpenAI and Microsoft in the NYT copyright case is consequential. But the stranger part is that Axios says the filing caught the Copyright Office and USPTO by surprise. AI competitiveness is a legitimate policy goal. www.axios.com/2026/09/19/j... #AI #Copyright #Journalism
Scoop: DOJ's copyright filing took key agencies by surprise
The department argues copyrighted works to train models should be considered fair use because that practice is new and transformative.
axios.com
Wikipedia has an unusual AI problem: Its knowledge is everywhere, while fewer humans are actually visiting the site. That matters because people don’t just consume Wikipedia. They fix it. They add sources. Correct errors. Update articles. Donate. www.adweek.com/media/wikipe... #Wikipedia #AI
Inside Wikipedia's Plan to Survive the Answer Engine Apocalypse
At Brandweek, the nonprofit’s director of brand explained why ubiquity without attribution is an existential threat, and how a new campaign aims to fix it.
adweek.com
The White House press pool isn’t a perk for whichever network happens to be on rotation. It’s shared infrastructure. www.cnn.com/2026/09/20/p... #PressFreedom #Journalism
What is the White House press pool, and why is it now under threat? | CNN Politics
It’s the most powerful office in the world, and hundreds of journalists want to cover it. But how?
cnn.com
The most important AI-agent question may not be “Are they autonomous?” It’s “Who pays for the mess they create?” An agent can save one person 10 minutes by sending 500 emails, hammering a reservation system or automating outreach. www.404media.co/theres-a-100... #AIAgents #TechPolicy
There’s a 100% Chance AI Agents Are Already Ruining the Internet
“AI agents” now have enough power and permission to be extremely annoying online.
404media.co
AI can answer a lot of simple questions now. That makes the next layer of journalism more important: What does this mean for me? What matters most? What can wait? What should I do next? www.niemanlab.org/2026/09/prot... #Journalism #ServiceJournalism
Protein grams and retirement plans: Service journalism enters its next era
"It's not enough anymore just to give a headline or answer a simple question. AI and social have really taken that over."
niemanlab.org
CrowdTangle mattered because it let journalists and researchers see how narratives were moving across Facebook and Instagram. Then Meta killed it. That’s what makes Arbiter interesting. www.niemanlab.org/2026/09/two-... #Journalism #PlatformAccountability
Two years ago, Meta killed CrowdTangle. Can a new AI tool fill the void?
Arbiter uses large language models and AI agents to help journalists see the next harmful narrative on social media before it peaks.
niemanlab.org
I like this direction much more than another “AI detector.” Apple’s new Reference Image feature tries to prove that a photo was actually captured by a real iPhone camera — essentially creating a digital negative at the moment of capture. www.niemanlab.org/2026/09/appl... #Journalism #AI
Apple launches a new way to prove a photo was shot with an iPhone (not generated by AI)
At its annual iPhonepalooza today, Apple announced a new feature aimed at making it easier to know where an image came from or whether it's been edited — the sort of provenance data that could be usef...
niemanlab.org
The best place to catch an AI fake is before publication, not after it goes viral. Press Gazette’s new tool puts AI/deepfake checks directly inside the newsroom CMS — text, images, video and audio. That’s a useful direction. pressgazette.co.uk/news/press-g... #Journalism #AIEthics
Press Gazette launches newsroom tool to stop AI fakes before publication
Press Gazette launches Made by Humans, a new tool for publishers which aims to protect them from publishing misleading AI-generated content.
pressgazette.co.uk
Meta says Community Notes now produces more corrective content than its old fact-checking program. But volume isn’t the same as coverage. Indicator found that only about 3% of Instagram posts already debunked by AFP received a Community Note. indicator.media/p/meta-says-... #FactChecking
Meta says Community Notes is bigger than fact-checking. That's not the whole story
Crowdsourced fact-checking on Instagram covered just 3% of false posts debunked by one fact-checking website
indicator.media
The courts will decide whether OpenAI and Microsoft infringed the copyrights of The Seattle Times and Newsday. But the economic question is bigger: Who pays for the reporting that makes AI answers useful? www.theverge.com/ai-artificia... #Journalism #AI #LocalNews
Seattle Times and Newsday sue OpenAI and Microsoft for infringement
The Seattle Times and Newsday want AI models built on their works destroyed.
theverge.com