Amazon Web Services has revealed an update to its facial recognition software, Rekognition, that can detect a person’s fear.

This update was announced on Monday along with improvements to accuracy and functionality of its facial analysis feature that can identify gender, emotions and age range.

The company claims the software can already accurately read seven ’emotions’, but it has now added an eighth – the ability to spot fear.

However, some experts have pointed out that while there is scientific evidence that suggests there are correlations between facial expressions and emotions, the way they’re communicated across cultures and situations can vary dramatically.

“Today, we are launching accuracy and functionality improvements to our face analysis features,” the tech giant said. “Face analysis generates metadata about detected faces in the form of gender, age range, emotions, attributes such as ‘Smile’, face pose, face image quality and face landmarks.

“With this release, we have further improved the accuracy of gender identification. In addition, we have improved accuracy for emotion detection (for all 7 emotions: ‘Happy’, ‘Sad’, ‘Angry’, ‘Surprised’, ‘Disgusted’, ‘Calm’ and ‘Confused’) and added a new emotion: ‘Fear'”

The ethical use of facial recognition, and its accuracy, particularly when deployed on a crowd, has caused concern throughout the world. From the London Met Police’s use that resulted in zero arrests and a 98% failure rate, to San Francisco’s outright ban of the technology, it’s now more famous for its problems than its benefits.

The UK’s Information Commissioner (ICO) has announced an investigation into the privacy aspect of facial recognition, which came to light this week after the owner of a development site in King’s Cross confirmed the technology was being used.

Amazon’s own Rekognition has also been a source of controversy after it was revealed that US law enforcement used the technology. There was even reports that AWS tried to offer the software to the Immigration and Customs Enforcement organisation, which sparked protests from Amazon staff.

Slack has revealed a set of features for admins that make it easier to manage organisations with large employee counts and busy channels.

The biggest change is that admins can now assign posting permissions more widely than a few select channels, and there is also a new set of APIs to automate the creation of workspaces with names, domains and descriptions.

“We believe this should be easier and today we’re introducing a couple of new features to do just that,” the company said in a blog post.

It comes a week after the company introduced more robust security measures for admins, including the ability to enforce data sharing limits and content blocks on certain devices, as well as a greater variety of two-factor authentication checks.

To start with, the announcement channels will create a single destination for the key information, so teams no longer have to decide what gets shared via email instead of what gets shared via Slack.

“We’ve long encouraged teams to send announcements in channels, where your employees are already working,” the company said. “To broadcast those updates clearly and without distraction, admins have always been able to limit posting permissions in the default ‘general’ channel. Now, for teams on our Plus or Enterprise Grid plans, we’re allowing users to set posting permissions for any channel.”

These come in the form of ‘granular’ controls which limit who can post in a channel and keep chatter to a minimum, leaving the space clear for the most important updates.

As for the new admin APIs, there will be a feature to invite thousands of members at a time, without the need to join the workspace themselves. Invite guest accounts to specific channels (including private ones), set a guest expiration date and customise a welcome message, delegate admin responsibilities to a specific member and automatically trigger the events above based on information collected via web forms.

“All these APIs work towards templated workspace creation and setup,” the company said. “In the future, admins can script the creation of new workspaces that will automatically be configured with their desired settings, content, apps and more.”

Senior executives from AWS and Google Cloud have hit out at Microsoft for changing how it charges customers using its software on other public clouds.

Some are even accusing Microsoft of trying to lock customers into a single vendor with a complex pricing structure.

From 1 October, Microsoft customers will have to pay additional fees if they want to run its software on AWS, Google or Alibaba cloud environments due to a change to its on-prem licences.

AWS CTO Werner Vogels took to Twitter to slam the change, saying: “Yet another bait+switch by $MSFT, eliminating license benefits to force MS use. 1st, MS took away BYOL SQL Server on RDS, now no Windows upgrades w/BYOL on#AWS. Hard to trust a co. who raises prices, eliminates benefits, + restricts freedom of choice.”

Google Cloud’s president, Robert Enslin, posted a tweet that suggested Microsoft was harking back to its old ways.

“Shelf-ware. Complex pricing. And now vendor lock-in. Microsoft is taking its greatest hits from the ’90s to the cloud,” he wrote.

AWS VP Sandy Carter said in a post that Microsoft was ‘awkwardly’ trying to force customers into Azure with the license change.

Carter said that Microsoft seemed to be taking from the “old guard software vendor playbook”. Firstly, by trying to put an end to Bring Your Own License (BYOL) for Windows Server purchased after October 1, 2019. She said that it would restrict customers ability to bring their own purchased licenses to their preferred cloud when using licenses purchased after the change comes into force.

She also accused Microsoft of trying to limit choice around SQL Server.

“If you are running SQL Server on the AWS cloud with Dedicated Host without software Assurance (SA) (which is allowed today) and want to upgrade to a newer version after October 1, you would be required to purchase a new SQL Server license with SA,” she explained.

Until now, Microsoft customers were allowed to use the same licence if they wanted to move a workload from an on-premise environment to a single-tenanted public cloud server.

“The emergence of dedicated hosted cloud services has blurred the line between traditional outsourcing and cloud services and has led to the use of on-premise licences on cloud services,” it said in a blog.

“As a result, we’re updating the outsourcing terms for Microsoft on-premise licences to clarify the distinction between on-premise/traditional outsourcing and cloud services and create more consistent licensing terms across multi-tenant and dedicated hosted cloud services.”

Customers now wanting to run Microsoft software on single-tenant cloud servers of AWS, Alibaba, Microsoft and Google will have to pay additional fees on top of the standard licensing.

Alastair Pooley, CIO at Snow Software, argues that the changes will almost certainly bring added complexity for customers.

“Microsoft’s recent change to licensing rules impacts the ability to “bring your own license” to dedicated cloud environments. If you are using the more common shared compute instances this will not affect you. Higher security or performance needs have led some companies to choose to be the only tenant on a physical machine and for those customers this will likely increase their costs and add complexity to their Microsoft licensing.

IT Pro has contacted Microsoft for comment

Contractors working for Microsoft are reportedly listening to personal Skype calls made using the app’s translation function. 

The video calling platform’s website does say that the company may analyse audio of translated phone calls in order to improve the service but it doesn’t state that this will be done by humans. However, through obtaining some Skype audio recordings and accounts of the human listening situation from unnamed Microsoft workers, Motherboard reported that the contractors are allegedly listening to personal conversations made through Skype. 

It is also reported that these contractors are reviewing recordings of voice commands made to Microsoft’s Cortana, which is currently a controversial trend with tech companies. Both Amazon and Google came under fire recently for reports that revealed workers were reviewing Alexa and Google Assistant recordings – Amazon later introduced a setting for users to disable human reviews.

The issue with it, however, isn’t that humans are listening to the device, but that it isn’t made clear in the terms and conditions. Skype Translator Ts&Cs state: “When you use Skype’s translation features, Skype collects and uses your conversation to help improve Microsoft products and services. To help the translation and speech recognition technology learn and grow, sentences and automatic transcripts are analysed and any corrections are entered into our system, to build more performant services.”

While it does state that it collects recordings to improve the service, it doesn’t explicitly say that these recordings are reviewed by humans contractors. Likewise, Microsoft’s own privacy statement also fails to make this crystal clear.

In July, after an unnamed Google worker leaked details to a Belgian public broadcaster about how the company reviews recordings made on its smart speaker, the tech giant responded in a blog post defending the practice.

As GDPR expert lawyer Frank Jennings told IT Pro at the time, speech recognition has progressed so far that we don’t expect humans to be involved at all, but there is still an obligation to be clear on the matter.

“While asking humans to assist with language recognition and booking fulfilment is a ‘legitimate purpose’ under GDPR, the real question is whether Google is doing so in a ‘transparent manner’ and for ‘specified and explicit purposes’,” said Jennings. 

While home speakers like Alexa and Google Assistant raise questions over privacy in the home, the reports of Skype Translator recordings present a potential worry for businesses.

“We strive to be transparent about our collection and use of voice data to ensure customers can make informed choices about when and how their voice data is used,” a Microsoft spokesperson said to IT Pro in an email. “Microsoft gets customers permission before collecting and using their voice data. We also put in place several procedures designed to prioritise users privacy before sharing this data with our vendors, including de-identifying data, requiring non-disclosure agreements with vendors and their employees, and requiring that vendors meet the high privacy standards set out in European law. We continue to review the way we handle voice data to ensure we make options as clear as possible to customers and provide strong privacy protections.” 

Which, again, still doesn’t explicitly say ‘humans’ are involved in this process.

Slack has introduced a slew of security features to give IT admins more control over which employees use can use the service and how.

These new features will help to implement limits on users and devices, including blocking both from accessing their company’s Slack account if they’re deemed to be suspicious or unsecured. 

The changes follow on from the company’s Enterprise Grid service, which was launched last year and promised more user efficiency and tighter security.

“Without proper controls in place, mobile applications can open your employees up to new security risks,” Slack wrote in a blog post. “To alleviate that, we’re rolling out new functionality to ensure that only the right people and approved devices can access your company’s information in Slack.”

To start, Slack is introducing new secondary authentication controls, allowing admins to implement additional layers of security in the form of Face ID, Touch ID, or generated passcodes. This also comes with a time limit function, after which users have to re-authenticate. There are also session management tools to remotely wipe a user’s mobile or desktop session in the event their device is lost or stolen.

Alongside these, Slack also unveiled data sharing protections. New domain whitelisting tools will be available for admins to control which workspaces can be accessed by its employees. Slack said this not only shores up sensitive company information, but it will also help teams focus on their immediate workloads. Another related feature  blocks users from downloading company information to an unmanaged device.

This is just the beginning, according to Slack. Session management controls will soon be added to the admin dashboards, which will allow them to define the maximum number of devices a single employee can be logged into at one time. What’s more, the company is working on a feature where admins can detect if a device has been jailbroken and then block its access to the app.

Slack said these new features are designed for IT professionals “who want to modernise and improve how their organisations work while maintaining compliance with their industry”.

For Jake Moore, cyber security specialist at ESET, it shows that security is slowly becoming important to the normal user, delivering what the people want rather than what the industry thinks the consumer needs.

“With Slack making great steps forward, adding more prominent security functions, it will hopefully make people more aware of the importance of authentication and other protection techniques,” he said. “It might even push other manufacturers into rolling out similar features as default.”

Spending on cloud services has grown almost 40% in the second quarter of 2019, with AWS and Microsoft Azure claiming half of the market.

Overall, the cloud market is heading for a world-wide revenue run rate of $100 billion per year, according to Synergy Research Group, with AWS taking up 33% of the that and Microsoft someway off in second with 16%.

AWS actually posted a slight slip in net sales with 37% growth compared to the same time last year. Despite going from $6.105 to $8.381 billion when compared to the second quarter of 2018, it’s the first time in over five years that its net sales have dropped below 40%.

However, not only is AWS still a large chunk of Amazon’s overall revenue (13%), it’s still bigger than the next four providers, Microsoft, Google, Alibaba and IBM, combined.

“When quarterly spend on cloud services is mapped out for the last twelve quarters, we are pretty much looking at a steep, straight-line growth profile,” said John Dinsdale, a chief analyst at Synergy.

“Amazon is maintaining its leadership position in the market, though growth at Microsoft is also noteworthy. In early 2016 Microsoft was less than a quarter the size of Amazon in this market, while today it is getting close to being half the size. These two cloud providers alone account for half of all money spent on cloud infrastructure services, which is impressive for such a high-growth, strategically important market.”

Microsoft’s public cloud computing platform, Azure, has firmly established itself as the second-place cloud provider, recently posting revenue growth of 63%.

Google Cloud, which holds 8% of the overall market, is generating $8 billion a year in run-rate according to parent company Alphabet’s latest earnings. The company also plans to invest in its sales force as it looks to close the gap on Microsoft and AWS.

IBM recently reported a drop in revenue, partly attributed to its acquisition of Red Hat but there is a suggestion that the open-sourced specialist is a big part of the IBM’s cloud strategy.

Alibaba, Salesforce, Oracle, Tencent and Rackspace made up the remaining market share with a combined 14%.

BT has announced a partnership with Canonical to develop and deploy its next-generation 5G core network.

The deal will see Canonical offer up its open-source virtual infrastructure manager (VIM) platform so that BT can run network applications as code and transition away from a hardware-based network to one that’s virtualised.

This open-sourced cloud-based approach will help BT to quickly deploy new services and allow it to stay ahead of the demand for 5G and Fibre to the Premises (FTTP), the company said.

Canonical operates its own distribution of OpenStack, a bundle of separate open source projects connected through APIs. When applied to BT’s own infrastructure, this will enable the separation of network hardware and software, turning core components into software applications so they can be updated faster and continuously integrated.

The result is that different network applications can share the same hardware across datacentres for a stronger and more resilient network. For BT, this means the speed at which its software can be updated could potentially lead to a new way of developing 5G services where it can build and deploy within a matter of weeks.

“BT has recognised the efficiency, flexibility and innovation afforded by an open architecture, and realises the value of such an approach in enabling its delivery of new 5G services,” said Mark Shuttleworth, CEO of Canonical. “We’re delighted to be working with them to deliver the foundation to this approach, which will underpin BT’s 5G strategy.”

In May, BT-owned EE switched on its 5G services, making London, Birmingham, Cardiff, Manchester, Edinburgh and Belfast the first places in the UK to experience the benefits of the next generation of mobile connectivity.

BT’s cloud-based full 5G core will be introduced from 2022 with its cloud-based architecture enabling ultra-reliable low latency communications and multi-gigabit-per-second speeds. The company said this phase of 5G will enable critical applications like the real-time traffic management of autonomous vehicles, massive sensor networks with millions of devices measuring air quality across the entire country and a “tactile internet”, where a sense of touch can be added to remote real-time interactions.

LinkedIn is set to migrate its computing workloads to its parent company’s Azure public cloud, some three years after being acquired by Microsoft.

The professional social network will shift all workloads from its own data centres to Microsoft’s cloud, with the project taking will ‘multiple years’ to complete, according to LinkedIn’s senior VP of engineering, Mohak Shroff.

Although Microsoft has largely left LinkedIn to its own management since it bought the platform in 2016, it has used a number of Azure technologies for improvements to content delivery, updates to its news feed and also to keep inappropriate content off the site.

The success of these improvements made Azure the right choice and one of the reasons for the decision to now move from its own datacentres, according to Shroff.

“Moving to Azure will give us access to a wide array of hardware and software innovations and unprecedented global scale,” he wrote in a blog post. “This will position us to focus on areas where we can deliver unique value to our members and customers. The cloud holds the future for us and we are confident that Azure is the right platform to build on for years to come.”

In October, Microsoft published details of how it had moved LinkedIn’s 14,000 employees off Google services and onto Office 365,  which took a couple of years. The move to Azure, however, may take much longer.

There are a number of Microsoft services that still don’t run on Azure, such as Office 365 and Xbox Live. Similarly, Amazon Web Services (AWS) has been trying to take its workloads off Oracle databases – most of which are Amazon systems set up prior to AWS.

Slack has unveiled a new version of its desktop app that’s reportedly faster and more efficient thanks to a complete rebuild of the platform’s underlying technology.

This will be rolled out to users in an update for Slack’s Windows and macOS desktop app. It promises big performance improvements, such as loading 33% faster and using 50% less RAM use than before, according to the company.

Over the course of 2018, Slack worked on shifting the web and desktop clients – which essentially use the same codebase – to a modern stack and away from jQuery and other technologies it used when first launched in 2012. The result is that Slack will no longer create a standalone copy for each workspace and take up RAM for each instance.

There’s a bit of risk-taking in this update as rewriting or changing the code of a platform can be extremely problematic. The desktop version of Slack is its oldest client, and as such, a few internal cracks were starting to show in the foundation, according to Slack engineers Mark Christian and Johnny Rodgers.

“Conventional wisdom states that rewrites are best avoided, but sometimes the benefits are too great to ignore,” the pair wrote in a blog post. “One of our primary metrics has been memory usage and the new version of Slack delivers…”

“These results have validated all of the work that we’ve put into this new version of Slack and we look forward to continuing to iterate and make it even better as time goes on,” the engineers continued. 

Part of the update included adopting React, a popular JavaScript, as Slack’s original user interface (UI) was built using HTML templates. These often need to be manually built whenever underlying data changed and the engineers said it was a “pain” to keep both data model and UI in sync.

The new update has kept the existing codebase but created a more modern section of it that the engineers have called “future-proof” and has been added incrementally. All UI components have been built with React and this code is both “multi-workspace aware” and also loads lazily, without adding unnecessary data – which is why the loading speed has improved.

Since launching, Slack has become the go-to comms app for startups and is often accused of enabling an ‘always-on’ work culture. Such is its popularity, the company recently went public and has become a key rival to some of the biggest tech companies around.

Currently, its main competitor is Microsoft Teams, which recently overtook Slack in terms of users, hitting 13 million – three million more than Slack.

IBM has said its acquisition of Red Hat will accelerate growth, despite reporting that current overall revenue has fallen for the fourth straight quarter.

Overall revenue was down 4.2% from the prior year to $19 billion, according to its second-quarter earnings, released on Wednesday. But the tech giant’s cloud business is on the up, with a 5% rise on $5.6 billion.

However, while these look positive, the figures still suggest IBM is far behind rival cloud providers AWS and Microsoft. Amazon’s cloud computing division leads the market and has continued to grow annually, with a massive 41% jump for the 1st quarter of 2019. While Microsoft’s Azure posted revenue growth of 73% in the third quarter of its fiscal year. Both companies will be reporting 2nd quarter earnings in the next few days and are expected to maintain their momentum.

According to IBM’s 2nd quarter figures, however, it is on track to achieve full-year expectations, especially if that excludes the recent acquisition of Red Hat – which was finalised last week.

“In the second quarter, we continued to grow in the high-value areas of the business, led by a strong performance across our Cloud and Cognitive Software segment,” said Ginni Rometty, IBM chairman, president and CEO.

“With the completion of our acquisition of Red Hat, we will provide the only true open hybrid multi-cloud platform in the industry, strengthening our leadership position and uniquely helping clients succeed in chapter 2 of their digital reinventions.”

IBM ended the second quarter with $46.4 billion of cash on hand, of which approximately $34 billion was used in July to close the acquisition of Red Hat. It does seem that the open-source specialist is a big part of IBM’s strategy to close the gap on its rivals, and lends credence to the idea that IBM’s fortunes are largely tied up in the success of Red Hat.

“On August 2, we will discuss how the acquisition of Red Hat will accelerate IBM’s revenue growth, contribute to our high-value model and enhance our free cash flow generation going forward,” said James Kavanaugh, IBM’s senior VP and CFO.