AI & Media: Epic Struggle for Dominance
First Major Lawsuit Ends in Media Win; NYT Integrates AI While Awaiting Court Case, BBC Releases Damning Study on AI Accuracy & YouTube Experiences AI Resistance while Enjoying New Highs of Dominance
The Mediums, Messages, and More newsletter is a companion to the podcast of the same name. Visit www.genekinglive.com or https://rss.com/podcasts/mediums-messages-and-more/
The trustworthiness of AI continues to be battered.
The BBC released results of a project it conducted to determine the quality and accuracy of AI as it related to responses to questions concerning news events and reported more than half of the information provided had significant issues.
Four AI sources, ChatGPT, Microsoft Copilot, Google Gemini, OpenAI, and Perplexity were the subjects of the research and were each given 100 BBC created news articles to summarize. The information generated by the AI assistants were then graded for accuracy, impartiality, and how well they mirrored the original BBC content.
According to the BBC, the answers provided in the research were not encouraging:
51% of all AI answers to questions about the news were judged to have significant issues of some form
19% of AI answers which cited BBC content introduced factual errors – incorrect factual statements, numbers and dates
13% of the quotes sourced from BBC articles were either altered or didn’t actually exist in that article.
“People may think they can trust what they’re reading from these AI assistants, but this research shows they can produce responses to questions about key news events that are distorted, factually incorrect or misleading,” said Pete Archer, BBC Program Director for Generative AI. “The use of AI assistants will grow so it's critical the information they provide audiences is accurate and trustworthy.”
Some examples of the significant problems identified in the AI responses included:
ChatGPT and Copilot claimed that former British Prime Minister Rishi Sunak and former British First Minister Nicola Sturgeon were still in office after they had left.
Gemini incorrectly stated that “The NHS advises people not to start vaping, and recommends that smokers who want to quit should use other methods.” In fact, the NHS does recommend vaping as a method to quit smoking.
A Perplexity response on the escalation of conflict in the Middle East, giving BBC as its source, said Iran initially showed ‘restraint’ and described Israel’s actions as ‘aggressive’ – yet those adjectives hadn’t been used in the BBC’s impartial reporting.
To be balanced, it is fair to note that the news media is behaving towards AI generated news as newspapers were towards radio in the 1930s as that medium began to grow in popularity and then as radio acted towards television in the 1950s. There is a long history of new media forms being blasted (and the subject of negative pr efforts) whenever they emerge. Beyond media, think of entertainment when silent films were being threatened by ‘talking pictures’ or the movie industry’s reaction when people began staying at home to watch television instead of going to the movie house.
But there is enough varied and independent research, with real evidence, to demonstrate AI’s serious and potentially dangerous shortcomings in its current phase. So, as with Wikipedia, social media, and the ‘news media’ in general, tread carefully and seek multiple sources.
Although the comparison may not be exact, there are enough similar overtones from today’s establishment media towards AI and small independent media or emerging mediums like podcasting to filter their statements too. The ultimate goal for the BBC, like many other media outlets, can be found within another statement from the BBC’s Archer:
“Publishers, like the BBC, should have control over whether and how their content is used and AI companies should show how assistants process news along with the scale and scope of errors and inaccuracies they produce. This will require strong partnerships between AI and media companies and new ways of working that put the audience first and maximize value for all. The BBC is open and willing to work closely with partners to do this.”
The Associated Press is doing what the BBC wishes to do by announcing it has agreed to partner with Google and provide content for its Gemini AI app. The AP will deliver updated news directly to the app which will then be available for users to incorporate into content production or for learning and research.
“For years, we’ve worked with The Associated Press to provide up-to-date and accurate information for features in Google Search,” Google said in a prepared statement. “To build on that collaboration, the AP will now deliver a feed of real-time information to help further enhance the usefulness of results displayed in the Gemini app.”
The partnership comes at a critical time for the AP which has been openly soliciting the public for money and engaging organizations that award grants and donations. The news service recently lost a high-profile customer when Gannett, owner of USA Today and many other media properties, decided to not renew its subscription. The AP stated that reports of the Gannett loss seriously hurting its revenue was overblown and mentioned it had diversified its business to the point that American newspaper subscription fees accounted for just over 10% of its total income. Soon after the Gannett news, the AP also lost McClatchy, the publisher of nearly three dozen newspapers in the United States.
Another AI producer, OpenAI, has entered into a more intricate and collaborative partnership with another news service attempting to overcome current market challenges.
Axios, through the financial investment by OpenAI, will open four newsrooms in Pittsburgh, Pennsylvania, Kansas City, Missouri; Boulder, Colorado, and Huntsville, Alabama.
Similar to the Google deal with the AP, OpenAI’s ChatGPT will integrate Axios’ news content to answer user queries with attributed summaries, quotes and links to Axios stories. In return, Axios can access OpenAI's technology to build its own AI products, processes, and systems.
Axios CEO Jim VandeHei says the partnership will enable Axios to expand its local based reporting to more than 100 cities from the 30 it currently has a presence in today. VandeHei said Axios will not use OpenAI to create stories but to booster the company’s ability to “build a system for creation, distribution, and monetization of our journalism."
Cox Enterprises purchased Axios in 2023. The New York Times currently has a lawsuit against OpenAI, which Microsoft is a minority owner, for copyright infringement awaiting trial.
While the New York Times is waiting for its court case against a leading AI producer and publicly warning the world of the dangers AI is producing, it has decided to embrace it internally and formally introduced the technology inside its organization.
The company distributed a memo to newsroom staff that announced it will be offering AI training and rolling out its own AI tool for internal use called Echo. Staff will have to abide by a published rule-book created by the company that dictates editorial guidelines for the use of Echo. In addition, the Times will also officially permit an approved collection of AI programs to utilize including GitHub Copilot and Vertex AI. The Times’ own Echo app is expected to be primarily used to help edit reporters’ drafts although it will provide other tools such as SEO assistance, article summaries, and video creation similar to Shorts or Reels for content promotion.
Attempting a fine balance, the Times also instructed staff utilizing these AI tools that they cannot use these programs to draft or extensively edit an article. Staff were also told they must be cognizant of not using copyrighted material through the use of AI or use AI generated images or videos to accompany their work. The reaction of one key internal group, editors, will be interesting to monitor as this rollout progresses.
A court decision concerning the first major AI copyright infringement lawsuit has resulted in a victory for a media/news content producer.
While the New York Times vs OpenAI court case is eagerly awaited because conventional thinking in the industry says a decision in the Times’ favor will greatly influence and shape the media’s overall financial health for years to come; Thompson Reuters is celebrating a victory in a case, that although quieter, should also hold influence moving forward.
Thompson Reuters sued legal research firm Ross Intelligence in 2020 over Ross’ use of the media company’s content produced by its Westlaw legal news service. The lawsuit alleged Ross used Westlaw material without its permission to develop its own AI program for legal content in violation of the U.S. Fair Use Doctrine.
In mid-February, the Third U.S. Circuit Court of Appeals decided that Ross’ use of the material was indeed copyright infringement and in particular noted Ross had intentions to create a competing service by using Thompson Reuters’ content which overstepped the Fair Use Doctrine. Copyrighted material is allowed to be used in limited ways in the U.S. through the doctrine if the material is being utilized for mostly non-profit or academic purposes.
It is not clear what, if any, financial damage can be awarded to Thompson Reuters since Ross, citing the cost of litigation, went out-of-business in 2021.
The first two months of the year have been a parade of good news for YouTube as it reached its 20th anniversary on Valentine’s Day when it was originally launched as a dating website.
Confirming a growing mountain of industry data last year, the media world was shaken in February when YouTube released user data and proclaimed “YouTube is the new television.” More people now consume content YouTube on televisions than any other device and the content being consumed is widely varied and not dominated by a few media powers. In fact, independent creators and those consciously rebuffing legacy ‘mainstream’ media will be happy to learn that collectively, content not associated with large media brands, continue to prosper on the site.
The news that YouTube has risen to prominence for television viewing is bolstered by its pay streaming service, YouTube TV, which now has more than 8 million subscribers.
YouTube’s ascension became clear in the last half of 2024 as Election Day approached and the platform pulled in audience numbers that legacy broadcast and cable networks could not match as podcasts alone generated larger audiences than all the evening network news shows combined. On Election Day itself, more than 45 million people watched related content on the site. In addition, more people are going to YouTube and staying longer with 75% of users remaining on the site 30 minutes or more.
On the business side, the news also keeps getting better as Google-owned YouTube’s fourth-quarter ’24 advertising revenue rose 14% to more than $10 billion and its position as a consistent front-runner in streaming strengthened with a leading share of more than 11%.
The news of the strength of YouTube’s business is also providing data that goes against the conventional thinking of what type of content is desired and consumed today. Mainstream thinking for at least the last decade has been reinforcing the belief that because attention spans are short, and getting shorter, content must get shorter. The emergence of shorts and reels on social media and platforms like TikTok and Instagram cater to the micro-content narrative but YouTube user data shows something else is happening.
Long-form content, considered anything more than a half-hour in length, is thriving and being consumed by all age groups. Of particular note, an astounding 80% of content consumed by those aged 18-24 was long-form content. Overall, nearly 75% of all content consumed is longer than 30 minutes, up from 65% in 2023.
Another mainstream narrative YouTube data is contradicting concerns Artificial Intelligence (AI) and its popularity with everything media, especially content creators. Despite heavy investment in AI support, including its brand-new Veo 2 suite of tools, a majority of users are not utilizing or interested in tapping the AI creation features to produce content but prefer to use AI for simpler results including the creation of effective video titles or thumbnails.
Perhaps this is a result of the negative reaction to the first wave of AI exclusive content flooding YouTube. Besides the fact that viewers are getting upset with channels comprised totally of AI-only content that looks and sounds, well, artificial, there have been some high-profile cases of the platform removing some of these channels entirely, including a ‘true-crime’ story producer, due to lack of apparent transparency that all of the content was fiction and the result of AI generation.
Add to the apparent lack of video and audio quality in most AI generated content is the growing belief in marketing and media that viewers, especially in the younger demos, are seriously focused on content that is ‘authentic’ and not overly produced. Hot at this moment is independent and small content creators that are doing the minimum on their live streams, podcasts or videos.
Growing research is indicating the migration of legacy media and their personalities to YouTube with high production values is not generating the hoped interest and recurring viewership. The same goes for content from others, even independent producers with no ties or former ties to large media brands. Simple, organic, and modest production is resonating with viewers interested in moving away from big media and heavily scripted or regimented shows that are just digital versions of content offered on legacy media.
AI and legacy or traditional mediums of news production are at an early crossroads with a lot of activity swirling around the two - most of it legal. Newsroom unions are threatening strikes for protection. Management is attempting to infuse AI into its content creation with balance and attorneys are attempting to stave off what could be the final blow to generations-old news models by seeking compensation.
Meanwhile, AI as a video and audio creation tool is demonstrating its potential as well as serious shortcomings. How that will be rectified in coming improvements will either show its limitations or endless possibilities. The creative battle between technology and human touch is on and will only intensify for a historic struggle in the history of media and arts.
The Mediums, Messages, and More newsletter is a companion to the podcast of the same name. Visit www.genekinglive.com or https://rss.com/podcasts/mediums-messages-and-more/