Governments are demanding information from tech firms at a growing rate, a study says that the Tik Tok algorithm is designed to make us more favourable towards China, the head of integration of RUST into the Linux kernel resigns and Canva raises its prices by as much as 300 percent claiming that its AI features make it worth it.
All this and more on the “The AI devil made me do it” edition of Hashtag Trending. I’m your host, Jim Love. Let’s get into it.
We’ve done a number of stories between this and our sister podcast Cyber Security Today that focus on the amount of data that companies are collecting on us from our internet and social media usage. We assume that most of that data goes to companies that want to use it to sell us products and services.
But it turns out that governments are also gathering data on us from these same companies. There is, according to a recent study from Surfshark, a cybersecurity company, an increasing number requests for user data from major tech companies.
That study shows the scale and scope of these requests over the past decade. From 2013 to 2022, government agencies worldwide requested data from nearly 9 million user accounts across Apple, Google, Meta (formerly Facebook), and Microsoft.
The United States leads the pack in data requests, with a staggering 3.3 million accounts requested over the decade. To make comparison easier, the report expresses the number in terms of requests per 100,000 people in the countries population.
And when we put it that way, the US government requests information on 989 accounts per 100,000 people. Or if you prefer, 1 in a hundred US residents .
That’s 9 times the global average.
Following the U.S. are Germany and the United Kingdom, with 850 and 453 accounts requested per 100,000 people, respectively.
The good news for Canadians is that we are number 27 on the list – with a mere 145 requests per 100,000 people or about 6 times less likely to request information.
Still, the trend is clearly on the rise. In 2022 alone, the report notes a 38% increase in requests compared to the previous year. In fact, the number of accounts requested has increased to more than 8 times the number of requests in 2013.
Now, how are tech companies responding to these requests? On average, they’re complying with 72% of them.
Remember the idea that Apple guards your information? Not according to his report. In fact, Apple leads the pack with an 83% compliance rate in 2022 but the others aren’t far behind. Google complies with 72.9% of requests, Meta about the same 72.8%, and Microsoft is the lowest, complying with 67% of requests.
As our lives become increasingly digital, governments are clearly seeing our online data as a valuable source of information for investigations. As we’ve said before, why bother spying on your population when you can just buy the data from data brokers – or in this case, just demand it from tech companies.
And given how easy it is to get that data, it’s a worthwhile question to ask – what are they doing with it and how are they ensuring it’s used only for the purposes they requested it for – you know, the same transparency that demand – or should demand – of the companies who gather our data.
The report and many useful graphs can be downloaded from the link on the show notes at technewsday.com or .ca
Source: Surfshank Study
Most of our focus on Tik Tok has been on the data they may be gathering on us. We might be more concerned, or at least equally concerned at the data they are sending TO us.
Researchers from Rutgers University and the Network Contagion Research Institute have found evidence suggesting that TikTok’s algorithms actively suppress content critical of the Chinese Communist Party, or CCP, while boosting pro-China propaganda.
How do they do it? Content Suppression: The study found that TikTok’s algorithms appear to systematically downplay sensitive discussions about issues like ethnic genocide and human rights abuses in China.
Keyword Analysis: Researchers examined search results for four keywords often associated with China’s human rights issues: “Uyghur,” “Xinjiang,” “Tibet,” and “Tiananmen.” They compared results across TikTok, Instagram, and YouTube.
Stark Contrasts: For example, only 2.3% of search results for “Xinjiang” on TikTok were considered “anti-China,” compared to 21.7% on YouTube and 17.3% on Instagram. Similarly, for “Tiananmen,” over 26% of TikTok results were “pro-China,” versus just 7.7% on YouTube.
And apparently, it’s working and changing user perceptions. A survey of American TikTok users revealed that heavy users of the app – those using it for more than three hours daily – showed a 49% increase in positive perceptions of the CCP’s human rights record compared to non-users.
This is purely a Tik Tok phenomenon. Similar effects were not observed with heavy users of YouTube or Instagram, suggesting this is a TikTok-specific phenomenon.
These findings raise serious questions about TikTok’s influence on public perception, especially among younger users. The researchers suggest this aligns with a strategic objective of shaping favorable perceptions of China among young audiences.
It’s worth noting that TikTok has challenged the study’s methodology, calling it a “non-peer reviewed, flawed experiment … clearly engineered to reach a false, predetermined conclusion.”
This study adds to ongoing concerns about TikTok’s relationship with the Chinese government. Both the Trump and Biden administrations have taken steps to address potential national security risks associated with the app.
As we navigate the complex landscape of social media and global politics, it’s crucial to remain aware of how algorithms might be shaping our perceptions of world events.
This has been your tech update. Back to you in the studio.
Sources: NBC News and multiple other sources
Wedson Almeida Filho, a software engineer at Microsoft who has overseen the Rust for Linux project, announced his resignation from the project last Thursday. This comes as a setback to efforts to incorporate Rust code into the Linux kernel.
Filho cited frustration with “nontechnical nonsense” as his primary reason for stepping down. He stated that after nearly four years, he lacks the energy to respond to certain types of interactions.
The Rust for Linux project aims to bring the memory safety advantages of the Rust language to the C-based Linux kernel. This effort began in 2020 and saw its first fruits with the release of Linux 6.1 in late 2022.
Memory safety bugs are a major source of serious software vulnerabilities in C and C++ projects. There’s been a push from major tech companies and government entities to adopt memory-safe languages like Rust.
But not all of the Linux community has embraced this with open arms. Filho’s resignation highlights ongoing resistance within the Linux kernel community to the adoption of Rust. He referenced a video where a respected kernel maintainer expressed reluctance to learn Rust.
Filho emphasized that no one is trying to force kernel developers to learn Rust or prevent refactoring of C code. These misconceptions seem to be a source of tension.
Despite the setback, Filho believes the effort to add Rust code to the kernel is mostly going well, though they’re encountering some roadblocks as new components are developed.
Some developers have expressed sympathy for Filho’s decision, noting the difficulties faced by Rust maintainers. Others, like Drew DeVault, have proposed alternative approaches, such as building a new Linux-compatible kernel from scratch in Rust.
This situation highlights the challenges of introducing significant changes to established open-source projects, especially one as critical and complex as the Linux kernel. It also underscores the ongoing debate in the tech community about the balance between maintaining existing codebases and adopting new, potentially safer technologies.
And it re-emphasizes how Open Source can be as much about community as it is about technology.
And Linus Torvald, the founder of Linux has remained silent during this great controversy. Which might be a good strategy.
Sources: The Register
If you’ve ever wondered how companies were going to make back the money they have put into moving AI into their features, we can only hope that Canva is not indicative of the general trend.
Canva a popular graphic design platform, which was seen as a simple and affordable alternative to the pricier Adobe suite, is set to dramatically increase prices for some of its users in 2024.:
Reports from various sources indicate Canva Teams subscribers could see their prices increase by over 300%. In the US, prices could jump from $120 per year for up to five users to $500 per year.
Canva attributes this increase to an “expanded product experience,” particularly citing the value added by their new generative AI tools.
Canva has aggressively rolled out AI features like Magic Media for text-to-image generation and Magic Expand for background extension.
The price increase may hit global customers even more heavily. In Australia, for example, a team of five users could see their annual costs rise from about $323 USD to $1,636 USD.
We haven’t yet gotten good data from Canada. We did read a report of one user reported seeing a price of $149.99 CAD for the year but who encountered a different amount of $380 CAD when attempting to purchase, which may relate to the Teams plan.
Canva has slid by price increases in the past. It changed its Teams subscription model in April, moving to $10 per month per user but then put on a minimum of three users. Existing users are now being moved to this structure.
And we’ve tried to validate some of these prices, but Canva does that thing we love to hate where you can get a free trial, you can see the features, but there’s no place to easily just find out what it costs.
What the impact will be on Canva, which, as we noted, was once seen as an affordable alternative to more expensive design software like Adobe’s suite – no one is certain. Predictably , the price increase has been met with criticism from users, with some announcing plans to cancel their subscriptions and switch to Adobe applications.
This price hike follows Canva’s acquisition of the company behind Affinity’s creative software suite and precedes a potential US public listing in 2026.
Canva is not the first to post a price increase for their AI features. Microsoft earlier started to charge 20 to 30 dollars per month for its Copilot Pro and Microsoft 365 Copilot.
While these amounts are not earth shattering, if this happens across all software programs, it could add significantly to the cost per employee for software and for companies with many users, it could give some IT management a real challenge at budget time.
Sources include: The Verge, Canva and others
Thanks for listening. I’m your host Jim Love, have a Wonderful Wednesday.
