​The BBC has opened the testing phase for its AI-powered voice assistant exclusively with those who routinely test updates to the Windows operating system.

UK members of the Windows Insider Programme can download the virtual assistant by grabbing the ‘Beeb BETA’ app from the Microsoft Store, where they can experiment with features and put the voice assistant through its paces.

These users will need to have installed the Windows 10 May 2020 update, however, which, Microsoft has incidentally warned its users against doing unless it’s shipped to their devices specifically. This is due to the developer identifying a string of issues that could lead to significant errors due to hardware compatibility issues.

The Beeb AI assistant aims to help users in the UK by navigating BBC services and access BBC content. There are no plans to build a device to rival Amazon Echo or Google Home, although Beeb will be built into the BBC website, iPlayer and all smart TVs.

The BBC’s voice team has previously collaborated with Microsoft to build the infrastructure behind Beeb with Azure AI services, as well as working with the industry giant on a number of other projects in the past.

“We’ve built Beeb because we know there is growing demand from people to access programmes and services with their voice – around one in five adults have a smart speaker in their home, and millions more have voice-activated devices in their pockets,” said BBC Design and Engineering’s chief operating officer Grace Boswood. 

“Much like the BBC did with iPlayer, we want to make sure everyone can benefit from this new technology, and bring people exciting new content, programmes and services in a trusted, easy-to-use way.”

The first version of Beeb will allow users to speak to access live TV and radio broadcast as well as programmes on-demand. Developers will add further features in the coming weeks and months, with feedback from Windows Insiders critical to the development path.

Not everything will be working as it should, the BBC added, with developers hoping feedback can shape what this early, stripped-down version will look like in the future when additions are made and compatibility with systems is expanded.

This beta-testing process will precede the rollout of Beeb in beta version to the general public for further testing and experimentation.

Dell Technologies has announced the launch of two new high-performance computing (HPC) solutions to help businesses run powerful artificial intelligence (AI) workloads in VMware environments and speed up digital transformation projects.

The new Dell EMC Ready Solutions are based on VMware Cloud Foundation and aim to help companies gain AI insights using the combination of Dell EMC systems and new features of VMware vSphere 7, including Bitfusion.

Tom Burns, senior VP of Integrated Products & Solutions at Dell Technologies, called AI a “game-changer” but added that the company’s customers are “lagging behind in adoption because they’re dealing with skills and infrastructure gaps”.

“We’re bringing together the power of Dell Technologies to help customers simplify the process of running AI workloads at scale in their familiar VMware environments,” he said.

The first Dell EMC Ready Solution is a GPU-as-a-Service (GPUaaS) designed to free up accelerator access by creating virtual graphics processing unit (GPU) pools. The system uses the latest VMware Cloud Foundation with VMware vSphere 7 support for Kubernetes and containerised applications in order to run AI workloads anywhere. The containers facilitate bringing cloud-native applications into production with the ability to move workloads as needed.

The second solution, for Virtualized HPC (vHPC), assists and economises the use of VMware environments for demanding HPC and AI applications in fields such as computational chemistry, bioinformatics and computer-aided engineering.

According to a recent study conducted by Forrester Consulting, Dell EMC Ready Solutions for vHPC is capable of delivering up to 18 times faster AI model development. It also delivers up to 20% faster hardware configuration and integration than self-installation and will provide an estimated return on investment of up to 111%.

Arthur Lewis, president of Server & Infrastructure Systems at Dell Technologies, explained in a blog post that “the Dell EMC Ready Solutions for AI: GPU-as-a-Service were designed for those who are just starting their AI journey or who already run smaller AI operations”, while “the Dell EMC Ready Solutions for vHPC were designed for those who want to advance their AI journey by virtualizing their high-performance computing workloads easily using the latest version of vSphere”.

The Dell EMC Ready Solutions for AI: GPUaaS and Dell EMC Ready Solutions for vHPC are available globally now, while the factory installation of VMware vSphere with BitFusion will only become available worldwide on Dell EMC PowerEdge servers next month.

A vulnerability in VMware’s Cloud Director platform, used by a host of cloud providers to manage cloud infrastructure, could allow attackers to gain access to sensitive data and seize control of infrastructure.

Rated CVSSV3 8.8, and assigned CVE-2020-3956, the code-injection vulnerability in the cloud service-delivery platform could allow an attacker to gain access to sensitive data and take over the control of private clouds within an enterprise.

Hackers could also exploit the vulnerability to gain control over all customers within the cloud. It also grants access to modify the login section of the entire infrastructure to capture the username and password of another customer, according to Citadelo, an ethical hacking company which discovered the vulnerability.

“In general, cloud infrastructure is considered relatively safe because different security layers are being implemented within its core, such as encryption, isolating of network traffic, or customer segmentations,” said Citadelo CEO Tomas Zatko.

“However, security vulnerabilities can be found in any type of application, including the Cloud providers themself.”

Citadelo was hired this year by a fortune 500 enterprise customer to perform a security audit and investigate their VMware Cloud Director-based cloud infrastructure. 

Using the code injection flaw, researchers with the company were able to view the content of the internal system database, including password hashes of any customers allocated to the information system.

From there, they were able to modify the system database to steal foreign virtual machines (VMs) assigned to different organisations within Cloud Director. The flaw also allowed them to escalate privileges from that of a customer account to a system administrator, with access to all cloud accounts.

Finally, they could read all sensitive data related to customers, like full names, email addresses or IP addresses.

The vulnerability was initially reported to VMware on 1 April, with patches released following towards the end of the month, and during May. Organisations that haven’t yet applied the fixes are still vulnerable.

Those affected include public cloud providers using VMware vCloud Director, private cloud providers using VMware vCloud Director, enterprises using VMware vCloud Director technology, and any government identities using VMware Cloud Director.

Zoom is planning to roll out stronger encryption for businesses and institutions that pay for its service.

Zoom’s security consultant Alex Stamos, who was poached by the company in early April, has confirmed the news but added that the plan was subject to change.

According to Reuters, Stamos has not yet decided whether stronger security measures could also potentially be rolled out for non-profit organisations or users in need of an extra layer of protection, such as political dissidents.

“At the same time that Zoom is trying to improve security, they are also significantly upgrading their trust and safety,” Stamos told The New York Times in an interview published yesterday.

“The CEO is looking at different arguments. The current plan is paid customers plus enterprise accounts where the company knows who they are.”

He added that providing full encryption for every meeting would leave Zoom’s trust and safety team unable to add itself as a participant in gatherings to tackle abuse in real-time.

Zoom hired former Facebook security chief Stamos following numerous security incidents which threatened the immense popularity of the video conferencing platform, such as ‘Zoom-bombing’, which led to numerous companies and institutions banning the use of the platform.

Zoom attracted millions of users and became the most popular video conferencing platform globally. With the majority of its audience using the free version of the platform, Zoom might be trying to increase the number of its paid users and regain the trust of businesses by promising a higher level of security for its paid subscription models.

However, a similar announcement from Facebook, which plans to implement end-to-end encryption across all of its messaging systems, has garnered criticism from its shareholders.

Zoom had been previously criticised for not using end-to-end encryption despite specifically stating that it does on its website. The company finally implemented the 256-bit AES-GCM encryption standard in late April.

​Microsoft is reportedly planning to replace contracted journalists that work on the homepages of MSN and its Edge browser with artificial intelligence. 

The tech giant won’t renew contracts for roughly 50 US employees at the end of the month, according to The Seattle Times, while The Guardian reports that around 27 journalists are being let go in the UK. 

The contracted employees, hired through staffing agencies, were reportedly notified last week that their services would not be needed beyond 30 June. The 27 UK employees affected, who were employed by PA Media, were told on Thursday that their jobs would be terminated as Microsoft is shifting away from humans in favour of automated news updates. 

A spokesperson for PA Media told The Guardian: “We are in the process of winding down the Microsoft team working at PA, and we are doing everything we can to support the individuals concerned. We are proud of the work we have done with Microsoft and know we delivered a high-quality service.”

Many publications have had to let journalists go due to the impact of COVID-19 and the outbreak has also accelerated a number of cloud-based technologies. According to some of those whose contracts are not being renewed, the curation of news on MSN is already partly automated. 

“It’s been semi-automated for a few months but now it’s full speed ahead,” one of the terminated contractors told The Seattle Times. “It’s demoralising to think machines can replace us but there you go.”

Microsoft is retaining full-time news producers, according to the reports, but the functions of the contracted employees will be taken up by automation. This includes using algorithms to sources trending news stories and optimising the content by rewriting the headline or adding different images or slide shows.

The contracted journalists also created news pieces, maintained editorial calendars of partner news sites and even assigned content to them but it isn’t clear if this will also be taken on by the software.  

“Like all companies, we evaluate our business on a regular basis,” a company spokesman said in a statement. “This can result in increased investment in some places and, from time to time, re-deployment in others. These decisions are not the result of the current pandemic.”

Welcome to the fifth issue of IT Pro 20/20, our sister title’s digital magazine that brings all of the previous month’s most important tech issues into clear view.

Each month, we will shine a spotlight on the content that we feel every IT professional should be aware of, only in a condensed version that can be read on the go, at a time that suits you.

To coincide with the second birthday of the General Data Protection Regulation, this month we take a look at the role of regulation in innovation. Rather than focus on GDPR principles and the importance of compliance, we thought it would be far more valuable to show how new rules are working to promote, and in some cases moderate, new technology and ways of thinking.

Our lead feature looks at the nature of corporate travel and how future regulations, as well as societal changes introduced as a result of the coronavirus pandemic, are likely to redefine what it means to travel for business. We also look at how authorities are attempting to rein in the development of cutting edge technology and whether it’s even possible to police something like an algorithm.

For those businesses confused about data laws in a post-Brexit UK, we’ve also put some of the most common issues to a panel of data protection lawyers to assess what the regulatory landscape might look like after the end of the transition period in January 2021.

As ever, you’ll also find a roundup of the four biggest stories of the month that are likely to reverberate throughout 2020.

The next IT Pro 20/20 will be available on Tuesday 30 June. Previous issues can be found here.

We hope you enjoy reading this month’s issue. If you would like to receive each issue in your inbox as they release, you can subscribe to our mailing list here.