It took years after acquiring Fitbit, but Google is finally shaking up its health and fitness products and services. Last month, the tech giant officially replaced the well-established Fitbit app with Google Health, and it launched the brand new Google Fitbit Air tracker with an ultra-minimal design that’s built for 24/7 use. I’ve tested them both for two full weeks, and here’s how they stack up.

Fitbit Air returns to the company’s roots in more ways than one. It doesn’t feature a screen like the Charge series. It doesn’t receive notifications like Versa smartwatches. It doesn’t come with any bells, whistles, or distractions. It’s a no-muss, no-fuss fitness band that tracks what Fitbit does best – steps, workouts, heart rate, oxygen levels, sleep.

While there’s a lot to like about Fitbit Air, there are a few negatives.

The device itself is a tiny pebble that houses the electronics, battery, and heart rate sensor. Its size alone is pretty impressive, considering the original Fitbit was about the size of a simple pedometer. The pebble fits into specially made straps meant to be worn on the wrist. It comes with the fabric Performance Loop band that is both soft and comfortable. You can also buy a secondary silicone Active band that’s great for sweaty workouts or polyurethane Elevated Modern band that’s meant to dress up the tracker when you go out.

Once it’s on the wrist, Fitbit Air is extremely lightweight. During my two-week test, I forgot I had it on half the time, which is exactly what you want from a device that’s meant to be worn 24/7. Despite its tiny weight and size, Fitbit Air can last approximately seven days between charges, though you may get a little more or less depending on how often you work out.

The most important part, though, is the data. How accurate is this tiny device? To compare, I wore Fitbit Air alongside my Apple Watch that has been on my wrist every day since 2015. Let’s see where they agree and how they differ.

Steps

During the test period, Apple Watch marked a higher daily step count 70% of the time while Fitbit Air was higher 30% of the time. The largest disparity left a 605-step gap (approximately a quarter of a mile) between devices at the end of the day, while they were only 15 steps apart on the closest day. There was a lot of variation between the two, making it difficult to decide which one was more accurate, so I resorted to a 100-step controlled test, where both devices accurately counted exactly 100 steps each. Ultimately, the difference between daily metrics likely boils down to the way both devices misinterpret slight hand movements — like typing on a keyboard all day — as steps.

Heart rate

Each device measures heart rate differently, with Fitbit Air logging data every several seconds and Apple Watch measuring heart rates every 4-6 minutes. This logging algorithm gives Fitbit Air more heart data to track over time, providing a clearer look at your heart health. For the most part, my Apple Watch and Fitbit Air agreed, with both devices crafting similar heart rate graphs each day. The only place where Fitbit Air falls short is during strenuous workouts. Sometimes, the heart rate sensor misses sudden rate spikes or lags behind by several seconds before it registers, potentially leading to inaccurate workout tracking.

RELATED: How the iPhone crushed young women's fertility

Godong/Universal Images Group/Getty Images

To end on a high note, though, Fitbit Air includes high/low heart rate notifications and irregular rhythm alerts to detect potential heart issues, which I thankfully didn’t get a chance to test during my review phase. Overall, Fitbit Air’s heart rate performance and safety features are impressive for its size.

Oxygen

Unlike Apple Watch, which lets you take an oxygen reading on demand, Fitbit Air only measures oxygen passively while you sleep. This data can then be used to help identify possible air obstructions or conditions, like sleep apnea. Comparing the two, Apple Watch was usually 0.5%-1% points lower than Fitbit Air, which is a huge discrepancy in the sensitive world of pulse oximetry. To be fair, though, wrist-based oxygen measurements are rarely as accurate as finger-based devices. The most important thing for fitness bands is how consistent the measurements appear from night to night, and both devices highlighted similar data trends.

Sleep

As someone who rarely gets enough sleep, tracking my good nights against my bad is critical for balancing energy, work, and responsibilities during the day. Thankfully, I’m happy to say that Fitbit Air excels at sleep tracking. Compared to my Apple Watch, both devices usually agreed on when I fell asleep and woke up within minutes of each other. There were even a few times when I woke up in the middle of the night for an hour and went back to sleep, which Fitbit Air captured perfectly. The coolest part is that you don’t have to put Fitbit Air into sleep mode (like Apple Watch) or tell it when you’re lying down for bed. It simply looks for physiological cues within your set bedtime and logs sleep automatically as you drift away. If you want more insight into your sleep health, Fitbit Air is a great place to start.

Zach Laidlaw

So what's the catch?

While there’s a lot to like about Fitbit Air, there are a few negatives worth mentioning: It doesn’t have a GPS sensor like most premium Fitbit devices and watches, and it doesn’t come with an altimeter either. This means that you’ll need to carry your phone with you on outdoor walks or runs to map your journey and record elevation information; otherwise, your workout data may not be as accurate or informative. Depending on your activity level, though, this may not even be an issue for you.

By the end of the test period, I walked away (slight pun intended) very impressed with this little device. It excels in heart rate detection, sleep tracking, and long battery life, and it’s more than good enough when it comes to monitoring workouts and oxygen at night. At only $99, Fitbit Air is easy to recommend to anyone who wants to get fit or simply keep tabs on health — that is, if you don’t mind giving your health data over to Google for at least the life of the product. And when it's time for an upgrade — well, you know the deal.

Like most tech companies these days, Google wants as much of you as possible in its ecosystem for life. Fitbit Air makes it easy and tempting to say yes.

On June 2, 2026, the White House released an executive order on artificial intelligence called "Promoting Advanced Artificial Intelligence Innovation and Security." The administration was at pains to explain what the order was not. It was not a burden or the Biden administration’s “top-down regulatory approach.” It was not, the official fact sheet insisted, mandatory licensing or pre-clearance or permitting of any kind. The document spent considerable energy describing its own absence.

This is a familiar American style of governance: The regulation that will not say its name.

How much technical judgment should a republic outsource to its security bureaucracy?

However, the order’s responsibility is actually rather specific. Within 30 days of signing, it directs federal agencies to prioritize the cyber defense of their information systems and requires the Department of Homeland Security to issue binding operational directives expanding AI-enabled defensive tools to federal agencies, state and local governments, rural hospitals, community banks, and local utilities. Within 60 days, it creates a classified benchmarking process, run by the Treasury Department and the NSA, to determine when an AI model’s capabilities have crossed a threshold and become what the order calls a “covered frontier model.” Developers may then voluntarily submit their model for government assessment. The government then gets 30 days to work with it before the developer shares it with anyone else.

The order is best understood as the third movement in a policy sequence that began with the first Trump administration. In 2019, the president signed an executive order framing American AI leadership as essential to both economic and national security but also emphasizing public trust, civil liberties, and privacy. In 2023, the Biden administration’s Executive Order 14110 described AI as holding both “promise and peril” and attempted something like a comprehensive social contract with the technology: safety and security, but also workers’ rights, civil rights, bias mitigation, and fraud prevention. On January 20, 2025, the new Trump administration rescinded that order. The declared rationale was ideological contamination. The Biden approach was “burdensome” and encoded “engineered social agendas.” The new policy would instead pursue American AI dominance, free from such considerations.

Government by bottleneck

What has been constructed is a security compact between the federal government and a small number of frontier-model developers. The developer will give the government access to a model, and the government will evaluate it. Together, they will decide who the “trusted partners” are who receive it next. The criteria for that evaluation are classified. The benchmarks are classified. The threshold for “covered” status is classified. Google and Sam Altman expressed support. The Business Software Alliance praised the order’s “voluntary and phased approach,” as though a process administered by the NSA with nondisclosure expectations was simply an industry working group.

RELATED: How to make DuckDuckGo your search bar go-to

herstockart/Getty Images

The order names its intended beneficiaries as "rural hospitals, community banks, and local utilities." These institutions do not have cleared staff, government relationships, or the organizational bandwidth to absorb classified defensive intelligence. They are institutions named in the fact sheet because they are sympathetic. The institutions that will in fact operate inside the order’s core architecture are frontier developers and their vetted, trusted partners. Everyone else waits for whatever the clearinghouse sees fit to distribute, assuming they have the expertise to use it.

This is a recognizable form of 21st-century American governance: highly centralized technical judgment, thin public transparency, and broad downstream dependence. The polity is told that the infrastructure will be hardened. The decisive knowledge about how, by whom, and according to what criteria is held somewhere else.

The order speaks fluently about attack surfaces and remediation and covered models and patch distribution. In this language, every institution becomes a node. Hospitals, banks, utilities, and federal agencies are all nodes. They are surfaces of cyber vulnerability awaiting protection. The world is rendered as a network diagram, and the only question is whether the correct agencies have been directed to harden it.

What the order cannot conceal beneath its operational specificity is that this is a theory of governance as much as technology. The Atlantic Council, in criticizing the order, noted that classified criteria and delegated executive discretion create a serious accountability gap. That gap is the design.

Red tape on steroids

The order presents itself as the rejection of bureaucracy, yet is an elaborate bureaucratic instrument with deadlines and interagency consultations, directives, classifications, threshold determinations, and enforcement priorities. It relocates the machinery of governance away from NIST’s open and collaborative risk-management culture and toward the executive security apparatus: the NSA, the Treasury, the national cyber director. These institutions will now pass judgment about which AI systems are of concern.

One can stipulate that these institutions are serious, technically capable, and acting in good faith. One can stipulate that the cyber threats are real; the frontier labs’ own safety documentation already makes clear that the most capable models can automate sophisticated intrusions against hardened targets, discovering exploits and chaining vulnerabilities at a pace no human team can match. The question is how much technical judgment a republic should outsource to its security bureaucracy.

Administration officials' bet is that urgency, expertise, and the specter of Chinese technological rivalry will supply sufficient legitimacy. They may be right. Urgency has carried American policy unquestioned a long way before. The order calls itself the enemy of regulation. It is instead regulation’s more exclusive cousin, with all the power, a fraction of the accountability, and a much shorter guest list.

For years, AI elites like Sam Altman, Jensen Huang, and even Elon Musk touted a future in which AI stole all the jobs and humanity simply accepted a life of meaningless unemployment while receiving a meager allowance of universal basic income until our dying days. Something changed recently, though, and several of those same elites are suddenly backpedaling on their promises of the past. What changed? There are a few possibilities.

Seemingly all at once, the CEOs of the world’s leading AI platforms, particularly OpenAI and Anthropic, both reneged on their opinions on artificial intelligence in the workplace. Where AI was once prophetically decreed to replace everyone's jobs, now these bots are being positioned as tools to enhance human productivity instead.

But why? For what reason would the AI CEOs, who once plotted workplace domination, suddenly turn back on their greedy aspirations? Did they suddenly remember that humanity must somehow live on after all the jobs dry up? That their companies will lose money if consumers don’t exist to buy products and services? That it’s actually evil to force people into unemployment amid a hostile takeover of the entire economy?

Public sentiment around AI is at an all-time low, and it continues to bottom out.

Maybe. Or perhaps something is forcing their hands.

Four reasons the AI job apocalypse is finished

It’s IPO time

Both OpenAI and Anthropic are at pivotal points in their meteoric rise to ubiquity. Neither company is turning a profit, and as time drags on, venture capitalists, who will never get a return on investment with generative AI, are more likely to reduce or even pull their funding. That means AI companies looking to survive the impending bubble have to find funding elsewhere. The answer is to go public.

The two AI giants plan to launch IPOs this year, and they need strong public support to drive value. If the companies are perceived as harmful or even complicit in obliterating the workforce and killing the economy, their IPOs will tank. As a result, they have pulled back on the dystopian warnings of mass unemployment as they tidy up their reputations to portray benevolent corporations bent on helping humanity instead of hindering it.

Reality check

While the AI CEOs promised a workplace revolution on the backs of their LLMs, the real-world applications for these platforms have fallen short of expectations. In May, Starbucks retired its AI-powered inventory system, despite supposed “improved product availability in stores” ushered in by the service. Employees responded by praising the change, saying, in effect, thanks for discontinuing automatic counting! The thought behind it was great, but the execution was proving difficult.

Also in May, a Gartner study revealed that 80% of companies that replaced employees with AI did not see better returns. Meanwhile, companies that added AI to their workforce to enhance the productivity of existing employees without eliminations saw the strongest gains, highlighting the need for skilled employees to coexist alongside AI platforms.

RELATED: Google's new daily helper knows all about you. Just how creepy is it?

Google's new daily helper knows all about you. Just how creepy is it? Marina113/Getty Images

Lastly, some companies, like Meta, are learning the hard way that AI isn’t a replacement for human intellect. As we reported in early June, hackers tricked Meta’s AI customer support bot into changing the passwords on high-profile Instagram accounts with little security to stand in the way. This was a massive blunder for Meta — which recently laid off 8,000 employees in favor of AI — in what became the company’s largest account breach ever.

Public protests

Public sentiment around AI is at an all-time low, and it continues to bottom out as time goes on. Just last month, numerous videos surfaced of college graduates booing commencement speakers for merely bringing up AI. Young people looking to enter the workforce, where entry-level positions are among the first to dissolve in the AI race, seemingly appear to hate LLMs. Since this demographic is the future customer base for AI giants, OpenAI and Anthropic would be stupid to continue to ruin young people’s lives with more promises of job replacements.

Another point of contention among the people focuses around data centers. Not only do these massive buildings devour local energy, there are also growing reports that they generate loud noises that have caused some unsettling health effects, including headaches, dizziness, nausea, sleep disturbances, and more.

Hefty price tags

Finally, companies are learning that AI is expensive to run at scale. Microsoft, one of the leaders in the AI space, canceled its Claude Code licenses for employees just months after starting the program. Although no official reason was given, the high cost and volume of Claude tokens required for sophisticated projects is believed to be the culprit. At the same time, Uber’s chief operating officer cited concerns over the high cost of AI that made it difficult to justify. Even AI GPU maker NVIDIA admitted that human employees cost less than AI bots.

During a recent event, Sam Altman was asked about the sizeable AI costs for businesses. He feigned ignorance, stating that "the issue never came up” in the past when setting the prices for companies. "People were totally happy with the amount they were spending.” That appears to no longer be the case.