Search is the Next Battle Front in the Information Wars
Diane Chang / Nov 27, 2024Diane Chang is the founder of Invisible Fabric, a consultancy on issues at the intersection of technology, media, and society.
In the wake of the 2024 US elections, efforts to ensure the integrity of online information ecosystems are at a critical juncture. With social media platforms retreating from more active moderation of civic mis- and disinformation, the role of search engines in providing access to trustworthy information is increasingly vital. As generative AI transforms how users engage with search, platforms must navigate the political challenges that have emerged in the past few years. Fortunately, there are key lessons and best practices from the social media age that AI search firms can learn from and adopt to safeguard the flow of reliable news and information.
The zone is flooded
Among experts tracking online election rumors and misinformation since the start of the US campaign season this spring, the sudden abatement of election fraud claims across social media platforms by late on November 5th was telling. The University of Washington’s Center for an Informed Public reported in a November 8th blog post, “As noted in our two Election Day memos, election fraud speculation… more or less died down by the end of Election Day.” With President-elect Donald Trump's victory and Vice President Kamala Harris's concession in the presidential race, it would appear that the networked campaign to sow election distrust was no longer needed.
Yet the campaign to preserve the ability to traffic in online falsehoods about election fraud, even if it was not necessary in 2024, accomplished its goals. The MAGA movement’s strategy to “flood the zone with shit” – as Trump advisor Steven Bannon reportedly put it in 2018 – has largely succeeded in the social media sphere as platforms have retreated from efforts to contain it. Meta started de-emphasizing news and political content on Facebook, Instagram, and Threads in early 2024. After years of intimidation by Republicans via House committee hearings and a spate of state laws aimed at limiting how platforms can moderate content to protect users, the major social media platforms have largely scaled back policies on election misinformation. With X transformed into a pro-Trump propaganda platform by Elon Musk and now the bully pulpit of the presidency, Trump and his network of surrogates and influencers have guaranteed distribution for any mistruths and incendiary comments.
Since Internet users are less likely to come across high-quality news and civic information through the passive-consumption feeds on social platforms, distributing high-quality information via search engines has become even more critical for the integrity of the online information ecosystem. In his first term, President Trump made unfounded claims that Google had rigged search results against him. Accusations of Google Search bias against Trump circulated during the recent election cycle as well, boosted further by Elon Musk. Search companies can expect more of such accusations.
With generative AI search products on the rise, newer companies that have not yet had to contend with the politicization of information will need to learn how to tackle the challenges of the current environment. They will learn that political questions are sometimes not ancillary but core to questions of product design. By making proactive choices based on learnings from the information integrity battles fought on social platforms, search companies can retain greater control of product design choices that users will find valuable and that help uphold the integrity of the information ecosystem.
Recommendations for builders of search products
The following are recommendations for companies that build consumer search products, including products that help individual users access news and general information from across the internet.
- Ranking news & information sources. When ranking news and news-like sources, firms developing search products should prioritize those that contain accurate, trustworthy, and reliable information. These are the criteria of information integrity agreed upon by 34 countries in the Global Declaration on Information Integrity Online. During breaking news events, search results should prioritize authoritative sources. Many organizations provide useful criteria on source quality, such as Ad Fontes Media’s “News Value and Reliability” tiers, the Global Disinformation Index’s disinformation risk score for US sources that assesses both content as well as an outlet’s operations, and NewsGuard’s website rating system based on trusted journalistic practices. It is worth noting that these ratings systems are mostly designed for sources operating in Western media markets where there is a relatively large degree of press freedom. Companies whose search products are more widely available outside these markets should understand and adapt to localized contexts.
- Resist the “view from nowhere” trap. The “view from nowhere,” coined by philosopher Thomas Nagel and often invoked by NYU journalism professor and media critic Jay Rosen, refers to the practice among many mainstream news organizations of insisting that every issue, such as climate change or an insurrection, has two equally valid sides; and that presenting both sides is what constitutes good, “impartial” journalism, even though one side is not grounded in facts. This phenomenon enables entities with authoritarian intent to exploit the existing media ecosystem and present falsehoods as truth. New search product companies should understand these tactics and the recent history of how bad actors have used them to accuse both traditional media and online platforms of bias. This context will help resist creating false equivalencies when presenting news sources in search results and citations.
- Provide transparency and explainability. Search product companies should disclose to the public the factors that drive search results to prevent misinterpretation of results and accusations of bias. For instance, they should explain how their ranking algorithms differentiate high-quality news and information sources from those that don't follow fact-based journalistic practices, such as tabloids and websites that traffic in rumors and engagement bait. They should embed media literacy principles into the design of user experiences and interfaces, such as those recommended in Ofcom's Best Practice Design Principles for Media Literacy.
- Don't reinvent the wheel. Firms should proactively partner with newsrooms, civil society organizations, and researchers to combat disinformation and information operation tactics. There is a strong network of experts around the world who are experienced in partnering with online platforms to strengthen information integrity. Leveraging existing expertise instead of reinventing the wheel will allow new search companies to more efficiently direct their resources and stay ahead of any malicious campaigns.
- Investing in sustainable business models for news is also good for business for search. It's no secret that the news industry has long suffered significant financial challenges as the online economy has rapidly changed. In more recent years, Meta’s move away from recommending news on Facebook, Instagram, and Threads, and Google's changes to news recommendation pages and search algorithms, even preceding recently launched generative AI search summary modules, have caused referral traffic to news sites to nosedive, with detrimental hits to their revenue. However, search products are only so good as the trust and reliability that users have in them; foundational to that trust is the ability to continue providing high-quality, reliable information to users, whether directly via links or through AI-generated summaries. The health of search companies' business is, therefore, symbiotically linked to the health of a broad, diverse, and productive news media ecosystem that can continue to produce authoritative and reliable news. New search companies should explore business models that help foster independent news organizations.
The organized effort to discourage social media firms from policing falsehoods will almost certainly expand to the new generation of AI-generated search products. But if search companies prioritize transparency, resist false equivalencies, foster partnerships, and support sustainable models for news gathering, these firms can play an important role in creating a healthier information ecosystem. The stakes are high, and there will be blowback, but the opportunity to build the next generation of tools and foster a resilient and trustworthy environment is critical.
Related Reading
Authors
