Thoughts on Tech News of Note - 05-15-2026

• Googlebooks?
• Canada's Bill C-22
• Maybe Anthropic Didn't Cry Wolf

Googlebooks?
If you follow tech news at all, you have no doubt been baptized in skeptical seascapes of the future of Google's new device, both appropriately and improbably named "Googlebook". Yes, we all know that there is already a product named "Google Books" and nevertheless Google hasn't even optimized search yet to deal with the dual nature of the search term. The truth is that the name is a bit of a problem because you can't hear that critical space in speech. I'm not sure how Google overcomes this or if they even feel it is something to be overcome. Surely, they thought of this, right? The URL for Google Books is books.google.com but if they go with something like googlebooks.com or even googlebooks.google.com, they must know there will be confusion. I'm going to assume they consider this confusion to be a feature and not a bug since in the end the confusion will likely lead to a Google product one way or another. The fact that Google hasn't yet said that they are releasing a 1st party Googlebook may also come out of the high probability of additional confusion.

Nevertheless, I've written on the subject of AluminiumOS powered laptops in the past and I said they needed very specific things to be even remotely successful in the marketplace:

1. A solid application ecosystem
2. Better hardware support (i.e. better mouse and keyboard support and support for other USB peripherals)
3. Better/smarter marketing

Google is already showing us that it's still bad at #3 and I don't really expect this to get better unless/until they hire a very competent ad agency that actually understands technology. To be honest, as I type that sentence, I'm surprised a Silicon Valley personality hasn't already spun up a marketing company that's good at this, perhaps now bolstered by AI to accomplish all matter of consumer deception not previously possible with manual labor. I expect someone will capitalize on this obvious business idea very soon now that there is so much money out there that bored millionaire and billionaire developers and venture capitalists are willing to throw at any idea smelling remotely of AI.

We don't know much about #2 because Google didn't show us much of the UI or functionality beyond the idea of shaking the cursor to bring up Gemini for a type of right-click-on-heroin pop-up. Some enterprising individuals have installed builds of AluminiumOS onto devices to show us some of the current features and it looks a lot like the Android desktop with some aspects of ChromeOS settings bolted onto it. One screen I saw showing accessibility settings reminded me a lot of what is currently available in ChromeOS. This gives me a tiny shred of hope that useful ChromeOS settings will make their way to Googlebooks as Android is in desperate need of more options and settings to compete better with MacOS and Windows.

As for #1, we also don't know much about what Google is planning here but announcing that Adobe Premiere will be available for Android later this year gives me another microscopic sliver of hope that perhaps other apps that are sorely needed on Android will arrive eventually. Android still needs more creator apps to compete not only with MacOS and Windows, but to go up against iPads, which are home for so many apps that support the creative arts. Android needs a polished professional music production app like Ableton or Pro Tools; it needs better visual art apps like Procreate, Blender, and Davinci Resolve; and there is a dearth of really good music transcription apps like Dorico and Sibelius. It could also use a full-featured Office suite that isn't so bad that people would rather use web apps.

Perhaps we will learn more next week at Google I/O, but based on what we've seen so far, I'd say Google still has a lot to prove and a long way to go to win over techies, much less regular everyday people who are happy using whatever they have now.

Canada's Bill C-22
The official name of the bill is the "Lawful Access Act, 2026", and it is designed to remake the rules for how law enforcement and intelligence agencies in Canada access digital data. The first part of the bill deals with user information requests and is supposedly standard fare, but the second part is where all the drama is brewing. There are three big ideas in part two that have tech companies lathered up and not in a nice clean way. Here's a list that AI created for me of the main concerns:

• Ministerial orders: The Minister of Public Safety can issue secret (JPW: secret to whom exactly? I mean, someone needs to know, right?) orders that would require companies to build technical capabilities to allow the government to have quick access to user data.
• "Systemic Vulnerability" Loophole: The bill says companies can't be required to introduce vulnerabilities, but this term is not defined and that leaves it open to future definition by future legislation that can't currently be predicted.
• User Metadata Retention: The law requires companies to store user metadata, including location, for up to one year, allowing the government to track user's devices.

You can imagine that the big American tech companies aren't logging these demands as new user stories and then rushing off to put them in future sprints ASAP. Apple is on record for saying that they will "never" build back doors into their products. This is not new; we've heard this sentiment, if not in fact these exact words, from them before when it came time to face the European Union (EU) on similar issues. Meta is also strongly opposed to the bill and has said the bill would turn private companies into "permanent government surveillance tools". I'd argue that Meta is already a permanent government surveillance tool, but that's probably not fair. It's perhaps better to say Meta products are permanent Meta surveillance tools. Words do matter. Signal, another often outspoken tech company, has also said that they'd prefer to leave the Canadian market than comply with this bill. So, the responses are exactly what you would expect from this motley crew.

I don't know where this will end, but so far, the Canadian government response has been remarkably similar to that of other countries. That is to say, instead of aiming to understand the concerns of customers and companies, they've indicated that all of this is just a big misunderstanding and they're not looking to destroy encryption or otherwise create new backdoors. This may be true as government often doesn't understand the technical ramifications of the laws they introduce, but this story has been around and around so many times in so many places that it doesn't quite ring as heartfelt. Canada has said they just need to do more education on the safeguards of the bill and that furthermore, they aren't really the right people to determine the best way to grant access without breaking something.

Above all else, that last statement is true. Tech companies do know a lot more about breaking things, especially while moving fast in the opposite direction of where you want them to go.

Maybe Anthropic Didn't Cry Wolf
This week, Google's Threat Intelligence Group (GTIG) announced that it identified and intercepted a cyber-attack that was generated by AI. The attack was made against a popular open-source web-based system admin tool, but we have not been told which one. The flaw allowed the hackers to bypass two-factor authentication. Google says it is confident that the exploit was manufactured by AI because they identified hallucinated security scores and found formatting that is typical in AI training data. Because the attack was terminated before it could be executed widely, it does not appear that any users were negatively impacted. Google worked with the tool's maintainer to patch the vulnerability.

When Anthropic unveiled its Mythos model, they used a lot of language that upset some people because it felt like they were using FUD (fear, uncertainty, and doubt) to hike up the attractiveness of their product, which isn't even available to most companies, much less the general public. Mythos was so good at exploiting vulnerabilities that Anthropic said they simply couldn't release it to the public. While some people were mistrustful that the situation could be as dire as what Anthropic posited, there have nevertheless been reports since of many flaws that were identified and patched that had gone unnoticed for years despite going through audits and security tools to check for possible exploitable code. More will be known once the initial 90-day exploratory period for Project Glasswing (Anthropic's program to partner with high-profile companies to patch vulnerabilities) expires and Anthropic publishes their findings. OpenAI recently released their own Mythos competitor and gave access to a number of their partners as well. This idea of AI being the best hacker ever always had legs and now it seems those legs are walking. Too bad they're walking like the zombies in World War Z.

We don't know the origination of this particular attack in terms of what AI tool was used or the amount of time it took to prepare it, but we do know that no matter what, cyber-attacks are going to continue and with the help of AI, it's going to become more common and harder to protect against. I'd like to see what Anthropic and OpenAI's plans are for protecting all of us and not just some big companies. It seems like it's just a matter of time before attacks happen in places where we neglected to focus attention and inflict damage we're just not prepared to address.

We'll have to see if anything beneficial comes out of the 90-day report when it ends on July 8, 2026.