Software

AI and Facial Analysis Job Interviews

Posted on by Andy Miller

Reports of the first job interviews conducted in the UK using Artificial Intelligence and facial analysis technology have been met with mixed reactions.

The Software

The AI and facial analysis technology used for the interviews comes from US firm HireVue. The main products available from HireVue for interviewing are Pre-Employment Assessments and Video Interviewing.

For the Pre-Employment Assessments, the software uses AI, video game technology, and game-based and coding challenges to collect candidate insights related to work style, how the candidate works with people, and general cognitive ability. The Assessments are customisable to specific hiring objectives or ready to deploy based on pre-validated models. The data points are analysed by HireVue’s proprietary machine learning algorithms, and the insights gained are intended to enable businesses to save time and use recruitment resources more effectively by enabling businesses to quickly prioritise which candidates to shortlist for interviews.

The Video Interviewing product uses real-time evaluation tools and can assess around 25,000 data points in one interview.  During interviews, candidates are asked to answer pre-scripted questions with HireVue Live offering a real-time collaborative video interview that can involve a whole recruitment team. The benefits of on-demand video-based assessments, which can be conducted in less than 30 minutes, are that recruiters and managers don’t have to synchronize candidates and calendars, and can evaluate more candidates, thereby being able to spend their time deciding between the best candidates.

Who Is Using The Software?

According to HireVue, 700+ companies use the software (not all in the UK) including Vodafone, Urban Outfitters, Intel, Ikea, Hilton, Unilever, Singapore Airlines, JP Morgan and Goldman Sachs. It has been reported, however, that the technology has already been used for 100,000 interviews in the UK.

Concerns

Even though there are obvious on-demand expertise, time and cost savings for companies, and HireVue displays case studies from satisfied customers on its website, AI and facial analysis technology use in interviews has been met with criticism by privacy and rights groups.

For example, it has been reported that Big Brother Watch representatives have voiced concerns about the ethics of using this method, possible bias and discrimination (if the AI hasn’t been trained on a diverse-enough range of people), and that unconventional but still good potential candidates could fall foul of algorithms that can’t take account of the complexities of human speech, body language and expression.

Robot Interviewer

Back in March, it was reported that TNG and Furhat Robotics in Sweden have developed a social, unbiased recruitment robot called “Tengai” that can be used to conduct job interviews with human candidates. The basic robot was developed several years ago and looks like an internally projected human face on a white head sitting on top of a speaker (with camera and microphone built-in).  The robot is made with pre-built expressions and gestures as part of a pre-loaded OS which can be further customised to fit any character, and the HR-tech application software that Tengai uses means that it can conduct situation and skill-based interviews in a way that is as close as possible to a human interviewer. This includes using “hum”, nodding its head, and asking follow-up questions.

What Does This Mean For Your Business?

Like the Swedish Tengai robot Interviews, the HireVue Pre-Employment Assessment (and possibly the video) appear to be have been designed to be used at the early part of the recruitment process as a way of enabling big companies to quickly create a shortlist of candidates to focus on. As businesses become used to, and realise the value of outsourcing as a way of making better use of resources and buying in scalable and on-demand skills and resources, it appears that bigger companies are also willing to trust new technology to the point where they outsource expertise and human judgement in exchange for the promise of better, and more cost-effective recruitment management.

AI, facial recognition, and other related new technologies and algorithms are being trusted and adopted more by big businesses which also need to remember, for the benefit of themselves and their customers and job candidates that they need to make sure that bias is minimised, and that technology is unlikely to be able to pick up on every (potentially important) nuance of human behaviour and speech.  It should never be forgotten that we each have the most powerful, amazing and perceptive ‘computer’ available in the form of our own brain, and for vast amount of medium and small businesses that probably can’t afford or don’t want to use AI to choose recruits, experienced human interviewers can also make good recruitment decisions.

That said, as technology progresses, AI-based recruitment systems are likely to improve by gaining their own experience, and be augmented, and become more widely available and affordable to the point that they become a standard first challenge for job applicants in many situations.

Email Signature Legally Binding For Lawyer

Posted on by Andy Miller

A recent ruling by the High Court that an email containing an automated signature is legally binding proved costly to the lawyer who sent such an email on behalf of his client that included the wrong price for a land sale.

£25,000 Below

The unfortunate lawyer, Daniel Tear, who sent an email to another lawyer setting out the terms for an owner’s land/property sale (but with the sale price listed as £25,000 lower than the asking price) the ruling about his email signature at the County Court in Manchester proved to be very costly.

In the case, which related to a dispute over the sale of land near Lake Windermere listed as a “jetty/boat landing plot/mooring”, it has been reported that the land should have been offered for sale at the asking price of £200,000 but (according to published court documents) but Mr Tear’s email to the lawyer of those wishing to purchase the land specified a price of “ £175,000 (one hundred and seventy-five thousand pounds”.

The lawyer acting for the buyer accepted the deal, and despite Mr Tear later emailing all the parties to say the deal had not been finalised by email, the court ruling went against him and his client.

Why?

According to the published court documents which refer to matters related to certain sections of the Law of Property Act of 1989, Mr Tear’s auto-signature (using Microsoft Outlook) which appeared at the bottom of his email, accompanied by the words “Many Thanks” (which link the email’s contents to the signature) were enough to make the contents of the email’s agreement binding.

In a hearing which considered the many difficulties around an email footer possibly being treated as a sufficient act of signing the judge stated that he was “satisfied that Mr Tear signed the relevant email on behalf of the Defendant” and that “the Claimants are entitled to the order for specific performance that is sought”.

Mr Tear’s argument that the case fell under Section 2 (1) of the Law of Property Act of 1989 i.e. “The document incorporating the terms or, where contracts are exchanged, one of the documents incorporating them (but not necessarily the same one) must be signed by or on behalf of each party to the contract” was, therefore, not accepted by the court.

What Does This Mean For Your Business?

As with most legal matters, if you read the court documents (here: https://www.bailii.org/ew/cases/EWHC/Ch/2019/2462.html) there were many different considerations based around the case. One thing that businesses can take away from this case, however, is that if you create and add an email signature section to the footer of your Outlook emails, even though it is automatically added to each of your emails, it may still prove to be enough to legally bind you to the contents of the email, even though you may have made a mistake. It goes without saying, therefore, that businesses need to be very careful to check that prices and quotes emails to clients (where an email signature is included) are correct and that any terms are clearly stated.  This ruling could now and in future have implications for many businesses in disputes relating to the contents of business emails.

Windows Virtual Desktop Generally Available Now

Posted on by Andy Miller

Microsoft has announced that its Windows Virtual Desktop is now generally available worldwide on Azure and will include Windows 7 free Extended Security Updates for up to three years.

Windows Virtual Desktop

Windows Virtual Desktop from Microsoft, which was announced last September but has just been made generally available worldwide, is a Cloud-based ‘virtual’ version of Windows that can be accessed by employees from any device from anywhere, provides full multi-session, and is always up to date.  The Virtual Desktop has been designed with modern working practice in mind where not all employees sit in an office, use just one device or work from secure locations.

According to Microsoft, Windows Virtual Desktop is the only virtual desktop infrastructure (VDI) that can provide simplified management, multi-session Windows 10, optimizations for Office 365 ProPlus, as well as and support for Remote Desktop Services (RDS) environments.

The Virtual Desktop enables Windows desktops and apps to be deployed and scaled on Microsoft’s Azure portal in minutes, and it includes built-in security and compliance features.

Supported Transition to Windows 10

One key sweetener of the new service for those companies facing the end of support for their old Windows 7 deployments is that it offers free extended security updates for the Windows 7 virtual desktop including more support options for previous app versions while users transition to Windows 10.

Migrate

Microsoft is keen to emphasise that its Virtual Desktop can work with your current Remote Desktop Services (RDS), and can therefore easily be migrated on Azure.

Trust

Microsoft is also keen to emphasise that businesses can trust the new Windows Virtual Desktop not least because Microsoft invests more than USD $1 billion annually on cybersecurity research and development, employs 3,500+ security experts, and Azure has more compliance certifications than any other cloud provider.

What Does This Mean For Your Business?

With Virtual Desktop, Microsoft is hoping to capitalise on the fact that many businesses have workers in multiple locations with multiple devices who need to have convenient and secure access to a constantly updated version of their desktop.  Microsoft also knows that companies are getting more confident about moving more of their infrastructure to the Cloud, and want a secure, scalable ‘as-as-Service’ offering where they don’t need to worry about having the expertise in-house.

The easy migration aspect of the service and the offer of extended Windows 7 support may be of value to businesses looking to make a leveraged move forward to Windows 10 and may help Microsoft retain valuable business customers.

Worldwide Rollout of ‘Personal Vault’ OneDrive Security Features

Posted on by Andy Miller

Microsoft has announced that the ‘Personal Vault’ security features for its OneDrive storage service are now available worldwide on all OneDrive consumer accounts.

What Is Personal Vault?

Personal Vault is a protected area in OneDrive that can only be accessed with a strong authentication method or a second step of identity verification.  These methods include a fingerprint, face, PIN, or a code sent to the OneDrive user via email or SMS.

The idea of Personal Vault is to add another layer of protection to important files, photos, and videos e.g. copies of documents such as a passport, driver’s licence, or insurance information. Even though the new feature means that users must go through a verification process, Microsoft has stressed that it won’t slow users down and that they should still be able to quickly access their files on a PC, OneDrive.com or on their mobile device.

Protection Against Lost, Stolen, or Unauthorised Access

The Personal Vault security measures should mean that files are not being stored unprotected on a PC and have additional protection, even if the Windows 10 PC or mobile device is lost, stolen, or if someone gains access to it or to the user’s account.

Other Security Measures

In addition to the second layer of identity verification, Personal Vault also includes security measures such as :

  • Scan and shoot, which enables documents to be scanned or photos to be shot directly into the secure Personal Vault area rather than leaving them on a camera or unsecured device.
  • Automatic locking of the Personal Vault after a period of inactivity to protect against private files being left open accidentally.
  • BitLocker encryption on Windows 10 PCs, so that all Personal Vault files are synced to a BitLocker-encrypted area of the local hard drive.
  • Restricted sharing so that Personal Vault and shared items moved into Personal Vault can’t be shared.

Some Limitations

Personal Vault does come with some limitations. For example, users with OneDrive’s free or standalone 100GB storage plan can store up to three files in Personal Vault, and Office 365 subscribers can store as many files as they wish as long as this doesn’t exceed their normal storage limits.

What Does This Mean For Your Business?

For Microsoft Personal Vault, this is another step in its competition with its most popular competitor, Dropbox, which recently partnered with BetterCloud to help with it provide cutting-edge data protection and orchestration.

For businesses using OneDrive, these new security features should prove attractive, particularly when most businesses need safe, fast Cloud storage for mobile devices and work PCs, and need an easy, reliable and convenient way to store sensitive and personal files and data.

Tech Tip – How To Sign a PDF Without Printing It

Posted on by Andy Miller

If you need to sign PDFs and return them (e.g. as part of your sales or buying processes) there is a way to do it without having to go to the time and trouble of printing out the PDFs, signing them, scanning them, and then emailing the scans back.

To sign the PDF’s electronically using Adobe:

– Open Adobe Reader.

– Open the PDF file you want to sign.

– Select ‘Fill & Sign’ and use the tools to create your signature – ‘Add Text’, ‘Add Checkmark’, ‘Place Initials’, and ‘Place Signature’. Other tools there include ‘Send or Collect Signatures’ and ‘Work with Certificates’.

– Select ‘Place Signature’ and use the pop-up window to select how e.g. Select ‘Type my signature’.

– Type your signature.

– Under the Review Your Signature, choose a signature style.

– Click Accept.

This signature can now be placed anywhere you want on a PDF.

Deepfake Ransomware Threat Highlighted 

Posted on by Andy Miller

Multinational IT security company ‘Trend Micro’ has highlighted the future threat of cybercriminals making and posting or threatening to post malicious ‘deep fake’ videos online in order to cause damage to reputations and/or to extract ransoms from their target victims.

What Are Deepfake Videos?

Deep fake videos use deep learning technology and manipulated images of target individuals (found online), often celebrities, politicians, and other well-known people to create an embarrassing or scandalous video such as pornography or violent behaviour. The AI aspect of the technology means that even the facial expressions of those individuals featured in the video can be eerily accurate, and on first viewing, the videos can be very convincing.

An example of the power of deepfake videos can be seen on the Mojo top 10 (US) deep fake video compilation here: https://www.youtube.com/watch?v=-QvIX3cY4lc

Audio Too

Deepfake ‘ransomware’ can also involve using AI to manipulate audio in order to create a damaging or embarrassing recording of someone, or to mimic someone for fraud or extortion purposes.

A recent example was outlined in March this year, when a group of hackers were able to use AI software to mimic (create a deep fake) of an energy company CEO’s voice in order to successfully steal £201,000.

Little Fact-Checking

Rik Ferguson, VP of security research and Robert McArdle, director of forward-looking threat research at Trend Micro recently told delegates at Cloudsec 2019 that deepfake videos have the potential to be very effective not just because of their apparent accuracy, but also because we live in an age when few people carry out their own fact-checking.  This means that by simply uploading such a video, the damage to reputation and the public opinion of the person is done.

Scalable & Damaging

Two of the main threats of deepfake ransomware videos is that they are very flexible in terms of subject matter i.e. anyone can be targeted, from teenagers for bullying to politicians and celebrities for money, and they are a very scalable way for cybercriminals to launch potentially lucrative attacks.

Positive Use Too

It should be said that deepfakes don’t just have a negative purpose but can also be used to help filmmakers to reduce costs and speed up work, make humorous videos and advertisements, and even help in corporate training.

What Does This Mean For Your Business?

The speed at which AI is advancing has meant that deepfake videos are becoming more convincing, and more people have the resources and skills to make them.  This, coupled with the flexibility and scalability of the medium, and the fact that it is already being used for dishonest purposes means that it may soon become a real threat when used by cybercriminals e.g. to target specific business owners or members of staff.

In the wider environment, deepfake videos targeted at politicians in (state-sponsored) political campaigns could help to influence public opinion when voting which in turn could have an influence on the economic environment that businesses must operate in.

Report Says Public Cloud May Double In Just Four Years

Posted on by Andy Miller

The new cloud market report from the Synergy Research Group shows that cloud-associated markets, such as the public cloud, are growing at rates ranging from 10% to over 40% and the annual spending on the cloud may double in four years.

IaaS & PaaS Biggest Growth

Synergy’s half-yearly report shows that, across the seven key cloud service and infrastructure market segments, revenues for operator and vendors in the first half of 2019 exceeded $150 billion, which is a rise in growth of 24% from the first half of 2018.

The biggest area of growth in the cloud infrastructure sector was in the infrastructure as a service (IaaS) and platform as a service (PaaS) market segments where there was a massive 44% growth rate.  IaaS is online, virtualised computing resources over the internet, and PaaS is where a provider hosts the hardware and software on its own infrastructure with PaaS products enabling developers to build custom applications online without having to worry about data serving, storage, and management.

The Synergy report also showed growth rates of enterprise SaaS at 27%, UCaaS at 23% and hosted private cloud infrastructure services at 20%.  The report also shows that spending on cloud services is now much greater than spending on supporting data centre infrastructure.

Infrastructure Investments

In the first half of 2019, cloud service provides spent $55 billion on the hardware and software used to build cloud infrastructure (evenly split between public and private clouds).  These infrastructure investments helped cloud service providers to generate over $90 billion in revenues from their cloud infrastructure services (IaaS, PaaS, hosted private cloud services) and enterprise SaaS.

Leaders

The Synergy report shows that the leaders in the IaaS and PaaS segments in the first half of 2019 are Microsoft, Amazon/AWS, Dell EMC, Cisco, HPE and Google.  Back in February, Amazon’s Web Services (AWS) reported a massive 45% growth in the revenue of the fourth quarter, mostly fuelled by big profits in its public cloud arm.

Other big names in that market segment include Salesforce, Adobe, VMware, IBM, Digital Realty, Equinix and Rackspace.

All these big players together account for over half of all cloud-related revenues.

What Does This Mean For Your Business?

The public cloud is being embraced by businesses as they seek to outsource and ditch traditional capital investment and maintenance problems and costs while reaping the benefits of having the pay-as-you-go scalability, security, and outsourced expertise that allows them to free up more of their own resources.  Cloud service providers are now investing heavily to win large slices of the cloud market with Amazon and Microsoft as market leaders, and as the Synergy report shows, this investment is delivering big revenues and impressive growth rates, particularly in the IaaS and PaaS market segments.

IBM To Offer Largest Quantum Computer Available For External Access Via Cloud

Posted on by Andy Miller

IBM has announced that it is opening a Quantum Computation Centre in New York which will bring the world’s largest fleet of quantum computing systems online, including the new 53-Qubit Quantum System for broad use in the cloud.

Largest Universal Quantum System For External Access

The new 53-quantum bit/qubit model is the 14th system that IBM offers, and IBM says that it is the single largest universal quantum system made available for external access in the industry, to date. This new system will (within one month) give its users the ability to run more complex entanglement and connectivity experiments.

IBM Q

It was back in March 2017 that IBM announced that it was about to offer a service called IBM Q that would be the first time that a universal quantum computer had been commercially available, giving access to (and use of) a powerful, universal quantum computer, via the cloud.

Since then, a fleet composed of five 20-qubit systems, one 14-qubit system, and four 5-qubit systems have been made available, and since 2016 IBM says that a global community of users have run more than 14 million experiments on their quantum computers through the cloud, leading to the publishing of more than 200 scientific papers.

Who?

Although most uses of quantum computers have been for isolated lab experiments, IBM is keen to make quantum computing widely available in the cloud to tens of thousands of users, thereby empowering what it calls “an emerging quantum community of educators, researchers, and software developers that share a passion for revolutionising computing”.

Why?

The hope is that by making quantum computing more widely available, it could lead to greater innovation, more scientific discoveries e.g. new medicines and materials, improvements in the optimisation of supply chains, and even better ways to model financial data leading to better investments which could have an important and positive knock-on effect in businesses and economies.

Partners

Some of the partners and clients that IBM says it has already worked with its quantum computers include:

  • J.P. Morgan Chase for ‘Option Pricing’ – a way to price financial options and portfolios. The method devised using the quantum computer has speeded things up dramatically so that financial analysts can now perform option pricing and risk analysis in near real-time.
  • Mitsubishi Chemical, Keio University and IBM, on a simulation related to reactions in lithium-air batteries which could lead to making more efficient batteries for mobile devices or automotive vehicles.

Quantum Risk?

Back in November 2018, however, security architect for Benelux at IBM, Christiane Peters, warned of the possible threat of commercially available quantum computers being used by criminals to try and crack encrypted business data.

As far back as 2015 in the US, the National Security Agency (NSA) warned that progress in quantum computing was at such a point that organisations should deploy encryption algorithms that can withstand such attacks from quantum computers.

The encryption algorithms that can stand up to attacks from quantum computers are known by several names including post-quantum cryptography / quantum-proof cryptography, and quantum-safe / quantum-resistant cryptographic (usually public-key) algorithms.

What Does This Mean For Your Business?

The ability to use a commercially available quantum computer via the cloud will give businesses and organisations an unprecedented opportunity to solve many of their most complex problems, develop new and innovative potentially industry-leading products and services and perhaps discover new, hitherto unthought-of business opportunities, all without needed to invest in hitherto prohibitively expensive hardware themselves. The 14 hugely powerful systems now available to the wider computing and business community could offer the chance to develop products that could provide a real competitive advantage in a much shorter amount of time and at much less cost than traditional computer architecture and R&D practices previously allowed.

As with AI, just as new technologies and innovative services can be used for good, their availability could also mean that in the wrong hands they could be used to pose a new threat that’s very difficult for most business to defend against. Quantum computing service providers, such as IBM, need to ensure that the relevant checks, monitoring and safeguards are in place to protect the wider business community and economy against a potentially new and powerful threat.

Autonomous AI Cyber Weapons Inevitable Says Security Research Expert

Posted on by Andy Miller

Speaking at a recent CloudSec event in London, Trend Micro’s vice-president of security research, Rik Ferguson said that AI cyberattacks operated autonomously are an inevitable threat that security professionals must adapt to tackling.

If Leveraged By Cybercriminals

Mr Ferguson said that when cybercriminals manage to leverage the power of AI, organisations may find themselves experiencing attacks that happen very quickly, contain malicious code, and can even adapt themselves to target specific people in an organisation e.g. impersonating senior company personnel in order to get payments authorised, pretending to be a penetration testing tool, or finding ways to motivate targeted persons to fall victim to a phishing scam.

AI Vs AI

Mr Ferguson suggested that the inevitability of cybercriminals developing autonomous AI-driven attack weapons means that it may be time to be thinking in a world of AI versus AI.

Example of Attack

One close example given by Ferguson is the Emojet Trojan.  This malware, which obtains financial information by injecting computer code into the networking stack of an infected Microsoft Windows computer, was introduced 5 years ago but has managed to adapt and cover its tracks even though it is not even AI-driven.

AI Launching Own Attacks Without Human Intervention

Theresa Payton, who was the first women to be a White House CIO (under president George W Bush) and is now CEO of security consultancy Fortalice, has been reported as saying that the advent of genuine AI has posed serious questions, that the cybersecurity industry is falling behind, and that we may even be facing a situation where AI will be able to launch its own attacks without human intervention.

Challenge

One challenge to responding effectively to AI cyber-attacks is likely to be that cybersecurity and law enforcement agencies must move at the speed of law, particularly where procedures must be followed to request help from and arrange coordination between foreign agencies.  The speed of the law, unfortunately, is likely to be much slower than the speed of an AI-powered attack.

What Does This Mean For Your Business?

It is a good thing for all businesses that the cybersecurity industry recognises the inevitability of AI-powered attacks, and although it fears that it risks falling behind, it is talking about the issue, taking it seriously, and looking at ways in which it needs to change in order to respond.

Adopting AI Vs AI thinking now may be a sensible way to help security professionals, and those in charge of national security to focus thinking and resources on finding ways to innovate and create their own AI-based detection and defensive systems and tools, and the necessary strategies and alliances in readiness for a new kind of attack.

AI Destined For McDonald’s Drive-Throughs

Posted on by Andy Miller

The acquisition of AI voice recognition start-up Apprente by the McDonalds franchise gives the restaurant chain its own Silicon Valley technology division and promises an automated ordering system for drive-throughs, self-order interfaces and the mobile app.

Apprente

Apprente is a Silicon Valley-based start-up (founded 2017, Mountain View, California) that specialises in making customer service chatbots.  Its acquisition by McDonald’s gives the restaurant chain its own AI-powered voice-based conversational system that can handle human-level interactions, thereby helping improve the speed and accuracy of orders.

It is thought that the Apprente system will not completely replace the traditional front of house staff, but may be used in mobile ordering or kiosks i.e. added to drive-through kiosks or sited nearby (and added to the mobile app) so that that food can be ordered by the customer’s voice, and transcripts of the order can be given to staff to ensure that the order is correct.  The transcript may also be presented or read to the customer when they pick the order up minutes later.  The technology may, therefore, provide time-saving, accuracy, and convenience benefits to both customers and staff.

Why?

There are a few key reasons why McDonald’s has gone down the tech route with its order taking.  These include:

  1. Competition from home delivery companies.
  2. 70 per cent of the company’s orders come through its drive-throughs but some reports show that McDonald’s may be relatively slow in getting its drive-through food orders out.  For example, a recent report (Oches’ 2019) shows that while the average wait in a Burger King drive-through is over 193 seconds, the waiting time in McDonald’s is considerably longer at 273 seconds.  McDonald’s ranked the tenth and slowest fast-food company in that report, but the addition of the voice-based conversational system could help speed things up.
  3. To give McDonald’s a technology development centre, the McD Tech Labs in Silicon Valley so that the restaurant chain can keep adding value through new technology and stay ahead in the market.

Other Acquisitions

McDonald’s has also recently acquired customer services personalisation company and AI start-up ‘Dynamic Yield’. With this deal, worth more than £240 million, McDonald’s can use the decision-logic technology to create drive-through menus tailored to its customers based on the time of the day, trends, previous choices and other factors.

What Does This Mean For Your Business?

For customers, the deployment of the new voice-recognition technology in addition to the Dynamic Yield (already deployed in 8,000 US drive-throughs) should make ordering food a faster and better experience.

For McDonald’s, the addition of the new technology and of a tech base in Silicon Valley to develop more of the same should help it to compete in a market that’s getting busier with companies that are using technology to reach customers and satisfy the same need for fast gratification.  The value-adding technology (combined with the fact that McDonald’s have a restaurant in most towns with a standardised and trusted product and brand) means that McDonald’s is taking steps to ensure that it stays ahead in a future where technology is an important competitive advantage in fast food delivery.   The new technology may also help McDonald’s address its current need to get orders ready more quickly and accurately while adding a novelty factor, talking point, and perceived advantage among customers.