Feed aggregator

Whether you get your forecast from an app on your phone, a website or a meteorologist on TV, most of the underlying information comes from the federal government.

(Image credit: Marcio Jose Sanchez)

Crypto exchange giant Coinbase is set to join the S&P 500 on Monday. It's the latest stunning development as Trump completely revamps the approach to crypto in the U.S.

(Image credit: Jon Cherry)

If you're trying to build muscle, getting enough protein is a must. But does it matter if that protein comes from meat or plants? A new study overturns assumptions.

Current and former officials say the current effort amounts to the dismantling of the civil rights division and its traditional mission.

(Image credit: Chip Somodevilla)

The pandemic helped supercharge a niche craft into a fine arts movement. Now the scene is more detailed and thoughtful than ever before.

(Image credit: Marina Totino)

Horse racing depends on thousands of workers without legal status, and industry leaders fear that Trump's soft touch toward the industry in his first term will not persist in his second.

(Image credit: Lydia Schweickart for NPR)

It's the latest big company to back away from its diversity pledges to keep the U.S. government happy.

(Image credit: Justin Sullivan)

Hindu temples offer prayers for a path to the U.S. But some in India were stunned by the way the U.S. deported Indians despite Prime Minister Narendra Modi's friendship with President Trump.

(Image credit: Diaa Hadid)

The MAHA movement loves the idea of getting rid of synthetic dyes in foods. But manufacturers face challenges that could drive prices higher.

(Image credit: BWFolsom/iStockphoto)

AI is undoubtedly one of the biggest developments to hit technology and business operations over the years. Tie that together with IT automation and everything suddenly appears a lot more complicated for business owners.

‘Agentic AI’ is one of the latest innovations to emerge recently. Otherwise known as AI Testing Agents, they use tools to help software programs run smoothly for companies. They can both monitor and test application performances to help identify any issues and optimize them for better performance. Early stage examples of AI agents minus the testing element include Siri, Alexa and Google Assistant.

In the past couple of years, I've observed firsthand how Artificial Intelligence is transforming entire industries, enhancing customer interactions, and refining operational efficiencies. It is these experiences that have led me to appreciate the nuanced benefits and challenges that come with integrating AI into business processes. Here, I share my perspectives on several key areas where I believe AI is making a significant impact.

The Pros and Cons of Using Test AI Agents in Business

From my point of view, the introduction of test AI agents in businesses heralds a new era of efficiency and capability. These agents excel in managing the inherent variability of AI responses, ensuring consistent performance monitoring amidst changing outputs. Their ability to handle large volumes of data through automated processes is particularly beneficial in environments where manual testing proves impractical.

Moreover, their role in facilitating continuous improvement is undeniable, as they provide the necessary feedback for iterative enhancements. However, these benefits are not without their challenges. The complexity of testing probabilistic AI systems requires sophisticated, statistical approaches that can be resource-intensive. Additionally, the rapid pace of AI evolution sometimes outstrips the development of standardized testing protocols, leading to potential inconsistencies.

Despite these hurdles, I believe that the strategic implementation of test AI agents is crucial for businesses in aiming to leverage AI effectively.

Implementing Agentic AI into the Workplace

For businesses, introducing Agentic AI into the workplace is a transformative process that I've seen yield enormous benefits when it comes to both efficiency and decision-making. In my experience, the key to successful integration is to start with clear objectives and thoughtful planning. This ensures AI agents are seamlessly incorporated into existing workflows, augmenting rather than disrupting established processes. Technical setup, including robust data pipelines and system integration, is what forms the backbone of this implementation.

Pilot testing is also crucial as it allows for real-world evaluation and adjustments. Moreover, ongoing training and support are essential to help employees adapt to and effectively interact with new AI tools. Continuous performance monitoring and iterative improvements based on feedback also play a critical role in realizing the full potential of AI agents in the workplace.

Industries Benefitting Most from Agentic AI Technology

In my view, several industries are currently reaping significant benefits from Agentic AI technology. Customer service has been transformed by AI agents capable of handling large volumes of inquiries in real time, significantly enhancing consumer satisfaction. Healthcare is another sector where AI has made a profound impact, assisting with diagnostics and patient management to streamline operations.

In finance, AI agents contribute to high-frequency trading, risk assessment, and fraud detection, improving accuracy and operational speed. E-commerce platforms utilize AI to personalize shopping experiences and optimize logistics, which enhances efficiency and customer engagement. These advancements underscore the versatile applications of AI across different sectors, driving improvements that are both transformative and beneficial for companies globally.

The Difference Between Conversational and Functional AI Agents

From my many years of experience in software, the distinction between conversational and functional AI agents is pivotal in tailoring AI applications to meet specific organizational needs. Conversational AI agents are designed for natural language interaction, making them ideal for roles that require engaging with customers or users, such as in service bots or virtual assistants.

On the other hand, functional AI agents are focused on executing specific tasks efficiently—whether it's data processing, automated testing, or controlling machinery. It is this specialization in functionality that allows businesses to deploy AI solutions that enhance productivity and accuracy in targeted operations.

Each of these facets of AI implementation and utilization reflects the broad spectrum of AI's capabilities and the diverse ways in which it can be harnessed to improve business processes and customer experiences. As we continue to explore and expand the boundaries of what AI can achieve, I remain optimistic about AI Testing Agents and their potential to drive further innovation and efficiency for businesses across the globe.

We feature the best AI chatbot for business.

This article was produced as part of TechRadarPro's Expert Insights channel where we feature the best and brightest minds in the technology industry today. The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/news/submit-your-story-to-techradar-pro

• 'Significant amount' of data stolen from Legal Aid Agency
• Up to 2.1 million records accessed, hacking group says
• Legal Aid online systems taken offline

The UK’s Ministry of Justice (MoJ) has revealed that a cyberattack on the Legal Aid system has led to the theft of a "significant amount" of data, including criminal records.

The MoJ was alerted to the attack on April 23 when data dating back as far as 2010 was accessed by the attackers.

Earlier this month, the MoJ said it was investigating a “security incident” and that payment information had potentially been accessed. The group responsible for the attack said they accessed 2.1 million pieces of data, but the MoJ has not confirmed this number.

‘Significant amount’ of data stolen

The MoJ said that the data accessed by the attackers “may have included contact details and addresses of applicants, their dates of birth, national ID numbers, criminal history, employment status and financial data such as contribution amounts, debts and payments.”

Legal Aid Agency chief executive Jane Harbottle issued an apology for the breach, adding that the breach “will be shocking and upsetting for people”. The MoJ is working with the UK’s National Crime Agency and the National Cyber Security Centre to secure systems, and the Information Commissioner has been notified.

The Legal Aid Agency is responsible for providing legal aid funding to over 2,000 providers. The amount administered in 2023/24 amounted to around £2.3 billion. The Agency’s online digital services have been taken offline as a result of the attack.

The MoJ has recommended that anyone who has applied for legal aid since 2010 to take steps to protect themselves, including increased vigilance against unknown phone calls and text messages, as well as updating or replacing weak or reused passwords.

"If you are in doubt about anyone you are communicating with online or over the phone you should verify their identity independently before providing any information to them," the ministry said.

Outside of personal data, it is likely that information relating to the barristers, solicitors, and other organizations, including not-for-profit organizations, was accessed by the hackers during the attack.

“The recent cyber attack on the Legal Aid Agency is yet another example of the real-world impact from digital vulnerabilities. When criminal records and other sensitive personal data are exposed, it is not just a matter of IT failure, it’s a breach of trust, privacy and even safety in this case. Many of the individuals affected may already be in vulnerable situations and could now face the added stress of not knowing where their data will end up or how it might be used," said Jake Moore, Global Cybersecurity Advisor, ESET.

“It highlights just how critical it is for public bodies and government agencies to invest in stronger cyber defences, quicker update times and better training as well as being transparent immediately when things go wrong. Delays in notifying victims or vague reassurances can often worsen the damage whether it’s a government agency or private company.”

Via BBC

You might also like

• Customer data stolen in M&S cyberattack
• Protect yourself with the best password managers
• Protect your family from online threats with the best parental control apps

With a war in Ukraine and the U.S. rethinking alliances, Britain and the European Union may need each other more than they thought. They're holding their first summit since Brexit Monday.

(Image credit: Leon Neal)

• "In response to EU regulatory pressure" Apple could be working to let users in Europe switch to third-party voice assistants over Siri
• The move comes amid indefinite delays to AI-powered Siri upgrades originally planned for March 31
• Executives are showing concern for the future of the company, as Apple falls behind competitors in the AI race

Apple is reportedly working on tweaking its iPhone operating system to allow European users to choose a third-party voice assistant as their default option over Siri.

The news comes via a bombshell Bloomberg Apple AI report by Mark Gurman and Drake Bennet, citing "a person with knowledge of the matter" and follows in the wake of Apple's public struggle to provide Siri with AI upgrades that were first revealed in June 2024.

It’s a situation that has already resulted in Apple pulling a range of AI-focused ads and "available now" claims from the Apple Intelligence webpage. These upgrades were originally supposed to launch in iOS 18.4 on March 31, but now appear to have been delayed with no firm release date.

This report also claims that Google Gemini is on track to be added in iOS 19 as a ChatGPT alternative, according to people with knowledge of the plan.

Alongside struggling to make Siri smarter, Apple is also now seemingly facing regulatory sanctions in the EU after already being fined for forcing users to use its own App Store.

Right now, if you ask Siri on an Apple Intelligence-enabled iPhone (iPhone 15 Pro or newer) something like "How does CRISPR DNA editing work?" the virtual assistant currently has no answer – except to ask users if they'd like ChatGPT to answer.

According to an Apple exec cited by Bloomberg, Apple's usual methodology is "we’re late, we have over a billion users, we’re going to grind it out, and we’re going to beat everyone", but that's not working this time around.

(Image credit: Apple) Where to now for Siri?

To back up its claims, Bloomberg’s report notes that Eddy Cue – Apple’s Senior Vice President for Services – is worried that "AI could do to Apple what the iPhone did to Nokia". Cue has recently been quoted during court testimony as saying it's possible the iPhone could be irrelevant 10 years from now, "as crazy as it sounds".

While it ultimately might not have a choice when it comes to changes in the EU, Apple would obviously prefer Siri to be able to compete with Gemini and ChatGPT head on, but at this point, it does not appear to have a clear path when it comes to closing the gap, with a senior member of Apple's AI team stating, “This is a crisis… It’s been sinking for a long time.”

While allowing users to ditch Siri for a third-party alternative would deal with the problem in the short term and could be a good thing for iPhone and Mac users, from Apple’s perspective it would also be an undesirable admittance of that current state of affairs.

(Image credit: Photo by PHILIPPE LOPEZ/AFP via Getty Images)

In his court testimony, Cue stated that for the first time in 22 years, Google searches on Apple devices fell last month. The provided reason was that users were turning away from search engines and towards AI tools for gathering information.

While Apple has a dedicated customer base that is going to continue to buy Apple devices almost no matter what, as AI becomes more integrated into people’s work and personal lives, it may be the deciding factor for less loyal users.

By comparison, Android devices have long allowed owners to switch to a different virtual assistant, something that’s continuing into the age of AI.

Apple has undeniably fallen behind when it comes to AI integration, but perhaps not disastrously so – and when it comes to such a world-changing technology, maybe that's not such a bad thing. Given how unreliable Google AI Overviews can be, it's certainly not an exact science.

Acknowledging that they need time to get Siri to a point where it can rival the others, and giving users further functionality in the meantime, seems like a win/win for Apple and its customers – if the iPhone maker can swallow its pride.

You might also like...

• 4 Gemini announcements I can't wait to hear at Google I/O this year
• The iPhone will soon turn 20, but the last thing it needs is an all-glass makeover
• Apple wants to connect thoughts to iPhone control – and there's a very good reason for it

The way security professionals and their businesses approach artificial intelligence is one of contrasts and contentions. We’re still in a state of flux as to whether AI will lead to true cyber safety, or leave us more vulnerable to increasingly sophisticated attacks.

Optimists view AI as a gleaming opportunity to improve cybersecurity tools at a rate that helps us stay ahead of attackers. AI can strengthen defenses at unprecedented speeds and deliver automated threat intelligence that guides both immediate and long-term decision-making.

A more pessimistic view however presents AI as a weapon for adversaries to launch increasingly sophisticated new and dangerous threats on unprepared individuals and organizations.

In my view, our reality sits somewhere in between.

What it means to be AI-secure

There are always two sides to a coin. Where AI is considered, risks can arise both internally and externally. For instance, hackers use AI’s speed and scalability to launch widespread cyber attacks from outside the business, while any mishandling of AI technology within can hand valuable assets directly over to them.

A sound AI security strategy considers both sides.

AI’s efficacy in helping businesses automate the data collection process is a well known advantage. However, organizations must not forget that speed and efficiency do not surpass security as a priority. For example, automating data collection with AI still requires anonymization when sensitive information is included, and it must be controlled so only approved individuals can access it.

Here, zero-trust and multi-factor authentication can help ensure only the correct individuals within an organization can access it. This practice stems from the tide of mishandling, improper storage and data leaks - purposeful or otherwise.

Like with any new software, businesses must also consider the security of third-party AI services. Are you certain that the AI model you are using is truly secure? Here, diligence must come before investment - organizations can have the strongest internal defenses only to be let down by poor third-party security.

We saw a recent example of third-party security failure by UnitedHealth, when an attacker gained server access through Change Healthcare, a third-party affiliate. Breaching this side door meant UnitedHealth’s highly protected customer data could be accessed in a matter of days.

There’s also the need to continually train AI models with the latest information. Making decisions on out-of-date data opens the door for misguided defense strategies. Before businesses make any AI-based decisions, they must remember to account for the recency and relevancy of the data they are processing.

What features to look for in AI-secure security platforms

While the global battle for AI superiority rages, businesses must carefully consider which AI tools will bring them the most security. The best providers of AI-ready security tools and services have preempted the concerns laid out above, and innovated their services to limit errors in managing an AI-ready world. Microsoft Security Copilot is a great example approach to AI-security and a cyber safe outlook.

Here are four features to look out for:

1. Real-time monitoring

When it comes to identifying vulnerabilities, monitoring threats and repelling attacks, cyber teams rely on powerful platforms to do the heavy lifting; there simply isn’t enough time or resources to do this on a case-by-case basis. Through proprietary AI, businesses gain transparent, comprehensive views of internal and external activity, network traffic, user journeys and data access.

From here, intuitive machine learning can recognize anomalies from within and outside the organization, harnessing historical data to analyze minute-by-minute threat environments and predict when attacks might occur and in what form.

2. Immediate action

Even the hardest working security professional can’t be in two places at once. Microsoft Security Copilot, among other systems, accounts for this, leveraging AI to monitor the constant stream of business activity while also intelligently flagging real-time threats.

Naturally, as incident response becomes more automated - although human oversight remains crucial - security teams can receive threat data almost instantly, with less manually intensive requirements, freeing up resources to immediately respond.

For the past two decades, we have been one step behind cyber attackers; with real-time response, we level the playing field.

3. Automation to reduce time and error

While there is something to be said for automated processes reducing human error, especially from time and resource-poor cyber specialists, it’s my opinion that this risk will always exist. At least with this mindset, we’re not falling into the old adage of ‘fail to plan, plan to fail’. Microsoft Security Copilot and other systems, for all of their advantages, are only as robust and foolproof as the security specialist using them.

Does automation decrease the chance of human error? Definitely. Does it cure it altogether? Definitely not. We must enable powerful new systems to remove manual task intensity, but they must remain directed by boots-on-the-ground specialists.

4. System cohesion

Businesses use many different systems for vital day-to-day operations. Data migration between separate systems, failing to train and educate employees, and poor security integrity across the tech stack, all increase the attack surface and therefore the organization's vulnerability.

The tech giants hold a significant advantage here. Take Microsoft Security Copilot, which has been built to integrate into Microsoft’s existing suite of tools, such as Azure Cloud, Defender, and Office products. This level of ‘baked-in’ security reduces the attack surface. It’s a strategic advantage that cannot be underestimated, offering a greater degree of monitoring, control and transparency.

Becoming AI-secure is the first step

The best piece of advice I and many of my colleagues can provide is to adopt a constant, deliberate and thorough analysis of AI opportunities and challenges.

Is there a new capability from your chosen provider that can help plug a long-known vulnerability to a greater extent? Invest and move quickly.

Have attackers launched a new AI-based threat that challenges existing defenses? It’s time to quickly analyze current strategies and find ways to innovate in turn, relying on your network to formulate the best industry-wide approach.

An AI-secure organization is one that commits itself to proactive cyber safety. AI can be our greatest tool in the fight to come, but we must not forget that attackers are thinking the same.

We've compiled a list of the best endpoint protection software.

This article was produced as part of TechRadarPro's Expert Insights channel where we feature the best and brightest minds in the technology industry today. The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/news/submit-your-story-to-techradar-pro

Sony recently launched one of its most-anticipated products of 2025 in the WH-1000XM6 headphones. The company’s previous efforts, namely the XM4 and XM5, have proven incredibly successful, so it’s unsurprising that Sony felt the need to continue the lineage.

Those previous models have actually proven to be so popular that both will remain on sale alongside their newly upgraded sibling.

I’ve been using the XM6 for about a week now to find out if they could dethrone my trusted Sonos Ace as my go-to over-ear cans. But, as you would have already guessed from the headline of this piece, they’ve fallen a bit short in that challenge for the top.

It’s got nothing to do with sound quality or noise cancellation, although I’ll touch on those features further down. Instead, it has everything to do with how they fit.

I’ve made no secret of the fact that I’m a fan of Sonos’ first attempt at producing a pair of headphones, predominantly because I find the Ace so darn comfortable, and that’s still true nearly a year after they launched in June 2024.

I’ve never worn the previous XM4 or the XM5 for prolonged periods, but having read my colleague Gerald Lynch's Sony WH-1000XM5 review, which praised the overall comfort level, I was expecting glorious things from the XM6.

According our new Sony WH-1000XM6 review, those expectations should have been met. For me, however, they've unfortunately missed the mark.

The different hinge designs of the Sony (left) and Sonos (right) headphones plays a big role in on-head comfort. (Image credit: Future / Max Langridge)

Perhaps I have a fussy head (hopefully not) or abnormally large ears (the shame) but, for me, the XM6 feel like they’re clamping down a little too much, at least when compared against the Sonos Ace. Sony has said it’s increased the clamping force of the new model compared to the XM5 to help aid with passive noise isolation, and it does indeed work on that front.

For my noggin, though, it’s just a little too forceful for comfort. They feel more akin to on-ear headphones when I wear them because the padding presses onto my ears a little too much. The Sonos Ace earcups, meanwhile, completely envelop my ears and feel much softer, and as a result feel way more comfortable.

I should add the realization that the Ace earcups envelop my ears more is an interesting one, because the openings are physically smaller than the Sony pair. I think the fact that the space inside the earcups of the Ace is deeper than on the XM6 is the reason for this peculiar phenomenon.

Despite weighing roughly 50g more than the Sony, I think the Sonos engineers have done a better job of distributing the weight of the Ace too. Again, I’m probably just being fussy, but it’s something I noticed as soon as I wore the XM6 for the first time and haven’t been able to shake.

In my opinion, one of the biggest factors to support my claim that the Sonos Ace feels more comfortable for me is the hinge mechanism used.

Both pairs feature hinges that integrate into the earcups, but the one on the Sonos Ace appears to have a greater range of flexibility, so it’s able to better conform to the shape of my head.

The Sony XM6 hinge integrated a folding mechanism for easier travel, but which I think limits its flexibility. (Image credit: Future / Max Langridge)

In contrast, the hinge on the Sony XM6 has a more limited range and so can’t adjust to the shape of my head as freely. That said, Sony's said the XM6 are inspired by an older hinge design, inherited from the XM4, that allow them to better fold for traveling. The Sonos mechanism, meanwhile, doesn’t fold away as compactly, so that could have something to do with it.

Don’t get me wrong, that extra compactness is absolutely a great feature to have – although one that’s admittedly a little lost on me, as I don’t travel that much – but my head is probably just a little too limiting. Perhaps the hinge will loosen up over time and the XM6 will fit me better? We’ll see.

Mind you, the carry case for the XM6 is, hands down, a winner over the Sonos design in my view. Sony has implemented a magnetic fastener this time around and it works a treat. The zip on the case of my Sonos Ace broke within a few months, rendering it next to useless – a trait I've noticed many other Sonos Ace owners have come across. I hope other headphone manufacturers follow Sony’s design choice with new pairs in the future.

The magnetic fastener on the Sony case (left) is a major design win in my book. (Image credit: Future / Max Langridge) One person’s music is another person’s sound

I’ve learned in my years writing about speakers, headphones and other hi-fi components, sound quality can be subjective and that, ultimately, if you like what you hear, then that’s the product for you.

There are instances where it’s impossible to say something sounds good. But more often than not, I’ve found that I don’t necessarily appreciate how a particular product sounds despite someone else thinking it’s great, or vice versa.

It’s a similar story here, because while our Sony WH-1000XM6 review heaps praise upon their sonic capabilities, I’m a little less enthusiastic. Or at least, I don’t think the Sonos Ace should be completely disregarded if your budget can afford either pair.

As I’ve said, though, sound is a personal preference and I actually prefer the Sennheiser Momentum 4 Wireless to either of these competitors for outright sound quality. However, comfort is still at the top of my wishlist, which is why I keep returning to the Sonos Ace. If I could get the Sennheiser sound in the Sonos body, it could just be my dream set of headphones.

I will admit the Sony headphones sound objectively better than the Sonos Ace. They’re more energetic, unearth a bit more detail and have tighter control over particularly tricky songs such as The Darkness’ I Believe in a Thing Called Love, resulting in a more cohesive listen.

While the Sonos Ace don’t sound bad in any sense of the word, I find myself having fun and bopping my head with the Ace in place when listening to dance music, which is my go-to genre – far more than when I’m listening to the same tracks on the XM6.

This could boil down to the fact that my ears don’t feel as constrained when using the Ace, resulting in a more 3D-like soundstage.

The Sonos Ace look and feel like a premium pair of headphones and the integrated hinge is excellent. (Image credit: Future / Max Langridge) Silence is golden

Now, when it comes to ANC performance of headphones, I personally don’t notice much of a difference between many of the latest pairs from the top brands. That could be because I listen to music quite loudly (perhaps to my own future detriment) and can’t hear any ambient sounds over it, including colleagues sitting next to me in the office. So it’s hard for me to judge if the noise-canceling tech is working as promised or not.

However, going off the in-depth reviews on TechRadar and speaking with my colleagues, the consensus is the ANC performance of the Sonos Ace isn’t class-leading. The opposite is the case for the Sony XM6, which are “the real deal as far as ANC is concerned” according to our reviewer.

This, too, might well be a personal preference. The amount of external noise you want to be blocked out could be different to my preference. If I can hear a train announcer telling me what station is coming up next, it doesn’t particularly bother me. If anything, it’s proven to be helpful if I’m taking a journey for the first time and I’m not sure where I’m going.

If you do want total silence, however, then I have to give it the Sony XM6 – though you could also consider the Bose QuietComfort Ultra Headphones, which have long been considered the class leader. The Sonos Ace are great, but ANC performance isn’t an area I can confidently defend them for.

Sony has redesigned the speaker driver for the XM6 headphones (left), which you can view when you take off the ear cushion. (Image credit: Future / Max Langridge)

With both the Sonos and Sony pairs costing a not-insignificant amount of money – Sonos Ace retails for $449 / £449 / AU$699 and Sony XM6 come in at $449 / £400 / AU$699 – neither should be an impulse buy. If you’re looking to spend this kind of money on a new set of headphones, then I'd strongly recommend auditioning both pairs first if you’re able to.

While you’re at it, you should probably give the recently released Bowers & Wilkins Px7 S3 a listen too, along with the aforementioned Bose QuietComfort Ultra Headphones. B&W’s premium over-ears have always been excellent where sound quality is concerned, but have often lacked in other areas such as ANC. That’s no longer the case with the Px7 S3, which have much-improved noise canceling capabilities. Combined with their gorgeous looks, they could be the ultimate pair in this price bracket.

The internal area of the ear cup of the Sonos Ace (right) is deeper compared to the XM6, which I find much more comfortable. (Image credit: Future / Max Langridge)

Therein lies a positive to take away from all of this: we now have so many quality choices when choosing a pair of headphones, that it’s hard for even experts like me to give a definitive answer on which brand one-ups the other. And given just how personal our choices are (one of my colleagues prefers the XM6 to the Ace for fit, for example), deciding upon and choosing the right pair is something only you can do for yourself.

You might also like

• 3 reasons I'd buy the Sony WH-1000XM6 and 3 reasons I wouldn't, based on my real-world testing
• View more of the best noise cancelling headphones
• I’ve tried Apple Music’s new Sound Theory playlists, and they’ve solved my sleep troubles

Republicans advanced their massive tax cut and border security package out of a key House committee during a rare Sunday night vote as deficit hawks allowed it to move forward.

(Image credit: Rod Lamkey)

Almost all of the wheat grown in the Pacific Northwest is for export, and even before President Trump's trade war, farmers were dealing with rock bottom prices and slagging global demand.

(Image credit: Kirk Siegler)

Hints and answers for the NYT Connections: Sports Edition puzzle, No. 238, for May 19.

Here are the answers for The New York Times Mini Crossword for May 19.

It may look "nutty" and like a "blender," but the designers say the craft could challenge helicopters