Hashtag Trending is brought you with the generous sponsorship of Zoho Canada. We thank them for making it possible for us to bring you this type of content. Zoho has a unique and powerful software suite to transform the way you work. It’s designed for businesses of all sizes and built by a company that values your privacy. Visit them at zoho. com.
Is your smart TV watching you? Meta’s told to pause the use of its AI models in Europe over the use of user data, Google’s new AI Gemini is caught parroting Chinese government propaganda, and Microsoft continues to backpeddle on the launch of its controversial Recall software.
These stories and more on this “I’ll be watching you” edition of Hashtag Trending. I’m your host Jim Love, let’s get into it.
An article on a site called Notus claims that any modern televisions come loaded with tech that tracks what viewers watch and how they watch it. And we consent to this tracking in those license agreements that we routinely say yes to without reading.
The biggest application of this tracking right now is in its us in political campaigns.
Data firms match viewing data to voter files, allowing campaigns to hit specific segments with tailored ads across platforms like Hulu or YouTube TV. During the 2020 election, one firm helped Biden’s team identify households getting bombarded with Trump ads on TV to counter with their own digital messaging.
Privacy advocates warn that voters have little transparency into how their television habits are being tracked and monetized for political gains. There are also minimal regulations governing how campaigns utilize this data compared to traditional TV rules.
As linear viewing declines, campaigns are rapidly shifting ad dollars to streaming to find elusive swing voters. Estimates suggest digital video ads will make up 40% of total spend this cycle, up from 27% in 2020.
While imperfect, the data-mining capabilities allow campaigns to surgically target and re-target voters across platforms in an effort to sway opinions. For better or worse, the living room screen is now a powerful portal into the modern campaign ecosystem.
Sources include: NOTUS.org
The Irish Data Protection Commission has instructed Meta to pause the launch of its AI models in Europe, following complaints regarding Meta’s plan to use user data to train its AI. Meta has agreed to suspend the launch of its AI models for the European market, at least temporarily.
This follows concerns around Meta’s plan to use personal data from Facebook and Instagram users to train its AI models without first obtaining consent. The complaint from NOYUA (for the protection of information) highlighted concerns about the use of user data without seeking consent first.
Meta has acknowledged this instruction from Ireland’s data protection authority, which leads on data protection issues in Europe. The decision came after complaints about using user data for AI training purposes without consent.
In blog post, Meta noted that they were “disappointed” at the last minute move as they felt they had incorporated legislative feedback and had informed the regulator of their plans back in March. The company further indicated that this was a blow to European innovation and competition in AI and that it would generate a substandard user experience compare to other areas which allowed the use of user data.
Source include: Reuters
Google’s new AI chatbot Gemini is being accused of being pro-Beijing. Testing by Voice of America found that when asked in Mandarin, Gemini refused to answer questions on issues like COVID protests in China, while repeating Chinese government talking points when describing leaders like Xi Jinping.
It referred to Xi as an “excellent leader” who will “lead the Chinese people continuously toward the great rejuvenation of the Chinese nation.” It said the Chinese Communist Party “represents the fundamental interest of the Chinese people and most controversially, it said the US government recognized China’s claim to Taiwan.
Experts believe the biased outputs likely stem from Gemini being trained on data flooded with Chinese state propaganda narratives from social media and websites. In contrast, Gemini gave more neutral responses to the same questions asked in English.
U.S. lawmakers have expressed concerns, warning the AI could be utilized to spread Beijing’s disinformation and erode truth. They are urging Google and other tech firms to increase transparency around their AI training data and methodologies.
The findings have reignited debates around potential national security risks of AI systems uncritically adopting authoritarian government stances. As large language models like Gemini become more prevalent, there are growing calls for building in safeguards against amplifying state propaganda and censorship.
This is not the first time Google has seen blowback on this topic. The company was reportedly developing a search engine for China that would meet the censorship restrictions of the Chinese government, but the company says it abandoned that project in 2018.
Sources include: Voice of America
Microsoft has backed down even further on its launch of Recall, the controversial recording software that captures a continuous replay of everything you do on your computer.
Originally slated for release to all Copilot+ owners on June 18th, Recall will now first go to members of the Windows Insider testing program in the coming weeks, with a public preview quote “coming soon” after that.
The delay follows weeks of pushback over privacy concerns with Recall’s ability to log sensitive user data like passwords and financial information. Microsoft had already been forced to make the tool opt-in instead of enabled by default.
On one hand we could see this as a way of Microsoft living up to its promise that it won’t rush software to market at the expense of security. The more cynical among us might point out that they seemed eager to move ahead until the massive negative publicity got their attention.
But with relatively few Windows Insiders using Copilot+ hardware, Recall may not get substantial real-world testing for months – slowing the rollout what they clearly hoped would be a key selling point as Microsoft faces rising competition in the AI PC space.
Sources include: The Register
And that’s our show for today.
Hashtag Trending goes to air 5 days a week with a daily news show, with a weekend interview show we call the Weekend Edition.
Show notes are at technewsday.ca or .com – either one works.
We love your comments. Contact me at editorial@technewsday.ca
I’m your host Jim Love, have a Marvelous Monday.