In the face of prevailing industry headwinds, many local news outlets around the country have shuttered. Their continued overreliance on an advertising-based revenue model, a vestige of the pre-internet times, leaves them vulnerable to further decline. Just in the last year, 130 more newspapers folded — a rate of two and a half a week — leading to a total loss of over 7,000 newsroom jobs. At the same time, others who took note of changing consumer preferences for the digital over print format saw a net increase of 105 outlets. To not just survive, but thrive in today’s fast-moving news media landscape, where consumers lean towards new formats such as short-form video, local outlets must embrace a tech-forward attitude. This means adopting the newest tools on the block, including AI, that advance newsroom productivity and capacity for producing high-quality, public interest journalism.

Building networks of trust among communities and sustaining vibrant information ecosystems are non-negotiables for healthy democracies. When local news deserts proliferate, communities experience diminished civic engagement: lower voter turnout, fewer contested races, and poorer public participation. It’s no secret either that the loss of reliable sources of information exacerbates urban-rural polarization. National news media rarely covers localities, and even less so when they are located in rural areas. Simultaneously, the most prolific outlets are also often the least attuned to the situation on the ground in most of working America, undermining valuable discourse in overlooked communities.

Reversing these trends requires an honest reckoning with the multi-faceted challenges that local newsrooms face. Their tenuous fiscal situations leave them under-resourced on several fronts. Local reporters must juggle multiple beats at a time, covering city council hearings, school board meetings, and small business openings, all the while performing serious investigative work. This workload makes devoting the requisite attention to unearthing and covering scoops well extremely difficult. Understaffing also inhibits other critical newsroom functions, such as fact-checking, translating, and distributing published stories. Without these supporting activities, local news risks its journalistic credibility and loses reach into communities. The shift to digital further exposes key skill deficiencies in local newsrooms, namely expertise in audience analytics and web design, that inhibit their ability to cater to relevant audiences. That’s especially important at a time when more young people are turning to social media for information that was once the bread and butter of a local newsroom — restaurant recommendations, classifieds, community events, and sporting fixtures.

Given the status quo, local newsrooms should welcome artificial intelligence as a tool for increasing high-quality coverage. AI holds enormous potential for local outlets because it excels at many functions that newsrooms currently lack due to budget cuts. For one, AI thrives when tasked with pattern recognition, which is highly useful for investigations that require journalists to process large datasets, such as troves of municipal documents. Another area of strength is AI’s aptitude for automating already standardized tasks like translation. For local outlets that serve multilingual audiences, accessible translations may meaningfully increase public engagement with existing coverage. 

AI may also help newsrooms scale their digital presence to better deliver content in the format that most audiences now prefer, including generating audio and video versions, summaries, and visual explainers. When local news media are able to engage well with their audiences online, their reach significantly expands. AI can help local newsrooms level the playing field in the face of upstart new competitors (like news creators on Instagram), and the larger hedge fund-owned conglomerates (who can invest in digital expertise and apportion the cost across multiple mastheads).

Indeed, it’s clear that newsroom productivity and the development of new capabilities are critical issues given new competitive dynamics and shifting consumer expectations. Consumers have a preference for new formats, as shown by the rise of TikTok as a source of news and the continuing long-term decline in consumers’ use of publisher-owned and operated platforms (e.g., websites and apps) in favor of social media. 

At the same time, due to staffing shortages, local reporters have a lot more on their plates than in the past. For example, local reporters covered an average of 3.8 different levels of government in 2000 (e.g., cities, counties, school boards, special districts, townships), which increased to 10 different levels of government in 2020. When they are already spread so thin, building trustworthy sourcing, chasing new leads, parsing through troves of documents, or otherwise pursuing time-consuming investigations becomes a near impossibility. 

With the acknowledgement that local newsrooms are under-resourced, AI can play a crucial role in expanding their investigative capacity. Tools like LocalLens, an AI-powered application launched in 2023 that automatically transcribes and summarizes local government meetings, allow journalists to cast their net far and wide for potential story ideas. LocalLens has also helped reporters connect with sources, such as student speakers at school board meetings, whom they might not have found on their own. Another example is the Associated Press’ 2024 launch of LocalLede, which uncovers relevant regulations from over 430 federal agencies for local jurisdictions. LocalLede can serve as a useful starting point for writing stories, flagging new federal announcements like changes to Earned Income Tax Credit (EITC) thresholds. Other AI tools have further enriched investigative journalism, helping reporters process large amounts of public records, including campaign finance disclosures, civil complaints, and municipal budgets.

AI tools also streamline repetitive tasks, allowing local news outlets to redirect limited resources to where they are needed most. Their pervasiveness in copy editing at major publications, for fact-checking and grammar and style corrections, should serve as a model for local newsrooms that still perform these tasks by hand. Natural language-processing-based services like Otter.ai save reporters countless hours whenever they need written transcriptions after conducting interviews. AI has also shown promise in automating translation work. When wildfires swept across Los Angeles in January 2025, the Boyle Heights Beat used a beta version of English-to-Spanish translation GPT to make their coverage and social media updates available to Spanish-speaking members of the community. 

For local newsrooms falling behind in their digital offerings due to the lack of technical expertise on staff, AI provides a means to catch up with larger counterparts. THE CITY, a non-profit outlet that covers all boroughs of New York City, currently offers AskNellie, an AI assistant that answers readers’ questions ranging from rent regulations to upcoming elections by connecting them with previous coverage. In 2024, the publication previously used artificial intelligence to create an online map of its areas of coverage, which provided transparency to audiences about whether or not they were sufficiently documenting underserved communities. ARLNow, a local Northern Virginia outlet, increased online engagement through publishing an automated daily newsletter, which they lacked the staff to compile before using AI to do so. 

While artificial intelligence has the potential to significantly improve local news media, it is important to approach its rollout with prudence. Failing to think through the specifics of applying AI to specific use cases will result in conspicuous misfires. For example, shortly after launching AI-authored high school sports coverage at several of its local news outlets, including the Columbus Dispatch in 2023, newspaper conglomerate Gannett pulled the project due to widespread backlash from readers for poor quality post-game summaries. Also in 2023, at CNET, editors had to issue numerous corrections to financial advice articles written by AI that contained obvious factual errors. Sports Illustrated showed poor judgment in November 2023 when the publication added fake journalist names and profiles to AI-generated content. When used judiciously, AI is a valuable tool for revitalizing local outlets, bringing them into the modern news media landscape, but it cannot be applied as a blanket solution without thoughtful consideration.

Ultimately, local news must keep up with the times. Embracing AI enables local outlets to make the most out of their limited resources — continuing to produce high-quality investigations, reach larger audiences, and put out competitive digital offerings. While local newsrooms may not completely reverse their decline anytime soon, adopting a tech-forward attitude will at least help them take steps in the right direction, ensuring that they can serve as more robust informational backbones for local communities who rely on their coverage.

Let me be clear upfront — I’m pro-AI. The technology has enormous potential to lower costs, increase productivity, and raise real wages. And once combined with future advances in robotics, the gains will be especially important in physical industries such as manufacturing and construction, which have both seen negative productivity growth over the past decade.

I also acknowledge the need to stay alert for problems, and to regulate as necessary.

But one problem that doesn’t overly worry me is the prospect of a massive short-term “extinction event” of jobs. For example, Dario Amodei, CEO of Anthropic, a leading AI company, recently told Axios that “AI could wipe out half of all entry-level white-collar jobs — and spike unemployment to 10-20% in the next one to five years.”

I find such a negative employment scenario for either college grads or non-college workers highly unlikely. First, previous forecasts of technology-driven job collapses have turned out to be premature, often signaling future job gains instead. It doesn’t make sense to just focus on job destruction from innovation without considering job creation as well.  

For example, the big ecommerce innovation was supposed to lead to a “retail apocalypse,” or a “retail meltdown,” eliminating millions of retail sales jobs and replacing them with fully automated websites. 

In fact, retail turned out to be fundamentally a logistics business, and having a good website was only the beginning of the e-commerce transformation. As I showed in a prophetic 2017 paper, getting fulfillment and delivery right was much more difficult and important, requiring more investment and more people. So while the number of poorly-paid retail sales positions declined by 15-20%, better-paid jobs doing ecommerce fulfillment increased even faster. Employment in the consumer distribution sector — including brick-and-mortar retail, fulfillment, and local delivery — actually rose by 1 million jobs from 2017 to 2024. 

Second, Amodei’s timeframe is far too short. The adoption of AI in business operations is likely to be costly and slow: Applications of AI to the existing workflow of a business will cut some costs by speeding up one step, while exposing other bottlenecks. It typically takes years to adopt a new technology across an entire business, and AI won’t be different.

In a pharma company, for example, speeding up the creation of marketing presentations using AI will do nothing to cut the cost and time of clinical trials. AI can help there as well — but it’s a much slower and painstaking process. Indeed, even if AI helps move us towards new treatments for conditions such as cancer and Alzheimer’s, the necessary lab research and clinical trials to validate the insights will themselves become job producers. 

Third, and related, what Americans complain about the most is the cost of necessities, such as housing, food, transportation, child care, and elder care. High tariffs may also increase the emphasis on domestic manufacturing. AI has a powerful opportunity to modernize these sectors, boosting capacity and creating a new wave of AI-complementary jobs, including for workers without a college education. 

Fourth, the educated employment market has been fed in recent years by a wave of Immigrants. Since 2019, fully one-third of the net new jobs for workers with a college degree or better went to foreign-born individuals. (That’s according to the published data from the BLS. However, adjusting for under-measurement of immigration in the post-pandemic period gives roughly the same percentage.) If that flow is choked off by Trump’s actions against foreign students and educated adult immigrants, we are more likely to end up with skilled labor shortages than surpluses.

There’s no denying that AI will have a big impact on the job market. But for most people, it could be good news rather than a job apocalypse. 

This past November, Canada’s Liberal government ordered TikTok to dissolve its Canadian business operations on November 6. The act does not remove the ability for Canadians to download or access the app, or post content, but presents a significant business disruption. TikTok has challenged this mandate, filing an application for judicial review with Canada’s federal court. 

This follows a more aggressive move by the U.S. Congress and the previous presidential administration to ban U.S. providers from providing hosting and other web services to TikTok, unless ByteDance divests from the app by January 19, 2025. Following a last-minute intervention by President Trump, TikTok was able to continue operating past the deadline but has still not returned to U.S. app stores and cannot be updated. Additionally, President Trump’s directive may be challenged in court and could fail before a more permanent solution is found. 

The similar moves by the U.S. and Canada follow similar justifications. TikTok is owned by ByteDance, a Beijing-based company. According to TikTok, 60% of the company is owned by global investors, 20 percent is owned by its founders, and 20% is owned by its employees. And while it may be owned by a Chinese-based company, TikTok itself is headquartered in Los Angeles and Singapore. But as a result of ByteDance’s association with China, governments across the world have expressed concerns that influence by the Chinese government, which yields wide unilateral authority to affect corporate operations in the country, could be used to force tweaks to its algorithm to advance subversive content or misuse sensitive user data. The data at TikTok’s disposal is vast: the app has become a global phenomenon and economic powerhouse, attracting over 1 billion users worldwide. Over 170 million Americans and 14 million Canadians use the app. 

This follows western relations with China growing increasingly tense in recent years. The West widely views the Chinese government as a foreign adversary, and American public opinion of China, for example, hit an all-time low in 2024. 

While moves to restrict TikTok may seem like low-hanging fruit for hawkish policymakers, the proposals to do so do not address the real concerns that apps like TikTok, but also social media apps broadly, present to user privacy and security.

First, while concerns over national security risks by a known foreign adversary can be legitimate, Canada, the U.S., and various other countries have yet to present definitive proof that TikTok has misused user data or made material platform changes at the behest of the Chinese government. At this time, the justification for the laws targeting TikTok has been speculation. 

As others have also pointed out, there is a broader concern in the TikTok debate about data privacy than just TikTok itself. TikTok, like many social media websites, collects and stores user data to reform its recommendation algorithms and ad targeting. However, as many have pointed out in opposition to the TikTok restrictions, this is not a unique feature of TikTok. All social media sites collect a plethora of user data for similar purposes. But if data privacy is the true concern, banning TikTok is only a small remedy to this issue that does not address the wider systemic need. Laws like Canada’s PIPEDA or California’s CCPA make progress on addressing this issue, but don’t go all the way to address the concerns laid out by data privacy activists.

Furthermore, TikTok is not uniquely vulnerable as a tool for the use of foreign interference. During the 2016 election, Facebook was used by Russian actors in an attempt to influence American public opinion on a large scale, and smaller attempts were made to use Google and Twitter for similar purposes. No evidence has been presented that TikTok has been used for foreign interference efforts.

TikTok serves as a lucrative platform for users, creators, and small businesses across the globe to build a virtual community. If users cannot access TikTok, they will simply move to alternative apps that can pose comparable or even more significant threats. We’ve already seen evidence of this movement in the wake of the U.S. TikTok ban. During the week leading up to January 19, U.S. downloads of RedNote — one of China’s most popular social media apps — nearly tripled, and over 700,000 new users joined the platform.

Outlawing TikTok in North America would have resounding economic and political consequences. According to an Oxford Economics report commissioned by TikTok, the app contributed over $24 billion to the U.S. GDP and $5.3 billion in tax revenue to the U.S. government in 2023. Presumably, a scaled impact on the Canadian economy and workforce would occur, too. If TikTok closes its operations in Vancouver and Toronto, hundreds of local jobs would be eliminated. Going after TikTok in North America not only has resounding economic repercussions — it is also an increasingly unpopular political position among voters. As of August 2024, only 32% of American voters supported a TikTok ban, falling from 50% in March 2023. 

Forcing TikTok to dissolve its operations or outright banning the app is a temporary solution to the broader data and national security issues nations across the globe are facing. Federal governments can maintain the economic benefits of TikTok and act in accordance with the political interest of their voters, while preserving national security and protecting user data by adopting comprehensive data privacy policies.