Video Labelling Causes Problems

Posted on by Andy Miller

Google has already been criticised by some for not calling out China over disinformation about Hong Kong, but despite disabling 210 YouTube channels with suspected Chinese state links, Google’s new move to label Hong Kong YouTube videos hasn’t gone down well.

Big Social Media Platforms Act

Facebook and Twitter recently announced that they have banned a number accounts on their platforms due to what the popular social media platforms are calling “coordinated influence operations”. In other words, Chinese state-sponsored communications designed to influence opinion (pro-Beijing viewpoints) and to spread disinformation.  Twitter and Facebook are both blocked in mainland China anyway by the country’s notorious firewall but both platforms can be accessed in Hong King and Twitter recently suspended over 900 accounts believed to originate in China. The reasons for the suspensions included spam, fake accounts and ban evasion.

Google Labels Videos

Google’s response, which some critics have seen as being late anyway has been to add information panels to videos on its Hong Kong-facing site saying whether the video has been uploaded by media organisations that receive government funding or public funding.  The panels, which are live in 10 regions, were intended to give viewers an insight into whether the videos are state-funded or not.

Problem

Unfortunately, Google did not consider the fact that some media receives government funding, but are editorially independent, and the labelling has effectively put them in the same category as media that purely spreads government information.

Google and China

Many commentators have noted an apparent reluctance by Google to distance itself from the more repressive side of the Chinese state.  For example, Google has been criticised for not publicly criticising China over the state’s disinformation campaign about the Hong Kong protests.  Also, Google was recently reported to have a secret plan (Project Dragonfly) to develop a censored search engine for the Chinese market and it’s been reported that Google has an A.I research division in China.

Disinformation By Bot? Not

There have been fears that just as bots can be a time and cost-saving way of writing and distributing information, they could also be used to write disinformation and could even reach the point soon where they are equal in ability to human writers.  For example, the text generator, built by the research firm OpenAI, has (until recently) been considered to be too dangerous to make (the ‘trained’ version) public because of the potential for abuse in terms of using it to write disinformation.  In tests (the BBC, AI experts, and a Sheffield University professor) however, it proved to be relatively ineffective at generating meaningful text from input headlines, although it did appear able to reflect news bias in its writing.

What Does This Mean For Your Business?

The influence via social media in the last US presidential election campaign and the UK referendum (with the help of Cambridge Analytica) brought the whole subject of disinformation into sharp focus, and the Chinese state media’s response to the Hong King demonstrations has given more fuel to the narrative coming from the current US administration (Huawei accusations and trade war) that China should be considered a threat.  Google’s apparent lack of public criticism of Chinese state media disinformation efforts is in contrast to the response of social media giants Facebook and Twitter, and this coupled with reports of the company trying to develop a censored search engine for China to allow it to get back into the market over there means that Google is likely to be scrutinised and criticised by US state voices.

It is difficult for many users of social media channels to spot bias and disinformation, and although Google may have tried to do the right thing by labelling videos, its failure to take account of the media structure in China has meant more criticism for Google.  As an advertising platform for businesses, Google needs to take care of its public image, and this kind of bad publicity is unlikely to help.

Audible’s ‘Captions’ Subtitles Feature Attracts Lawsuit From Publishers

Posted on by Andy Miller

The Amazon-owned producer of spoken audio entertainment ‘Audible’ is facing a lawsuit from the Association of American Publishers (AAP) on the grounds that its new “Audible Captions” speech-to-text subtitles feature may violate copyright law.

Audible Captions 

Audible Captions, which was announced by the largest producer of audiobooks via a YouTube video back in July, is a feature that allows text captions to appear on-screen and progressively highlights the words as a novel is narrated. The feature also highlights and gives definitions for certain words in the captions and allows the user to translate text into other languages.

Objections – Lawsuit 

Audible’s plans to roll-out the Captions feature attracted almost immediate complaints and concerns by authors, publishers and literary agents on social media over possible copyright law violations, along with accusations that Captions appears to make quite a few mistakes per book. Eventually, a lawsuit was filed at the District Court for the Southern District of New York by the Association of American Publishers (AAP) which includes seven of the top US publishing companies, such as Penguin Random House, and HarperCollins Publishers.

Injunction

The lawsuit, which seeks a preliminary injunction to stop the September launch of Audible Captions argues that the feature could give Audible a competitive advantage over other audio-book providers who aren’t in a position to utilise speech-to-text technology, and that displaying the text of Audible books may amount to illegal reproduction and distribution of those books, thereby potentially breaching copyright laws and adversely affecting publishers’ profits. The AAP members also appear to be angry that the mistakes (transcription errors) made by the AI aspect of Captions could add up to the equivalent of 18 pages of inaccuracies in 300-page book. The AAP’s legal action has also attracted the support of the US Authors Guild. Executive Director, Mary Rasenberger, makes the point that “Text and audio are different book markets, and Audible is licensed only for audio. It has chosen to use its market power to force publishers’ hands by proceeding without permission in clear violation of copyright in the titles.”

What Does Audible Say? 

Amazon-owned Audible has argued that Audible Captions are an educational and accessibility innovation, and that the Captions, which allow listeners to simply follow along with a few lines of machine-generated text as they listen to the audio are not and were never intended to be a book, and therefore, can’t be judged like one (with copyright law).

What Does This Mean For Your Business? 

In addition to their anger over allegedly not being consulted by Audible about using the feature, the big publishers and Authors Guild appear to see Captions as a competitive advantage that represents a threat to their existing benefits, profits, and market positions. For Amazon, a company that has grown and diversified and made major inroads into multiple markets, the lawsuit is not only another dose of bad publicity e.g. following recent concerns by China Labour Watch (CLW) about possible child labour being used in the manufacture of the Amazon Echo, but it’s a reminder that there are still other powerful players in the publishing market and that laws regarding copyright need to be studied and adhered to, no matter how big the market player. It is not clear when Captions will be released but it is unlikely that Amazon’s Audible would want to be delayed too long in releasing a value-adding feature that could provide a competitive advantage.

Grammar Correction Capabilities For Gmail and G Suite

Posted on by Andy Miller

Google has announced that it is rolling out new real-time, AI-powered spelling and grammar correction capabilities for G Suite users and personal Gmail accounts.

Real-Time, As-You-Type

The real-time, as-you-type spelling autocorrection is now available to all G Suite users and personal Gmail accounts, but the Grammar suggestion feature is only available to G Suite users.

Autocorrection Spelling

The new features use AI to spot and highlight mistakes with spelling and grammar, and even the use of tenses.

The autocorrection spelling feature uses coloured (red) squiggly underlines to highlight spelling mistakes, the mistakes can be auto-corrected, and the changes are temporarily underlined in grey so that you can see the difference between the two.

Grammar

The AI-powered grammar correction feature, which has previously been available in Google Docs, uses a squiggly blue line to highlight errors in grammar, and pop-ups appear to help make changes to the grammar and allow the user to mark them as correct or switch them back to how they were.

Default

The autocorrect spell-check feature has been added as a default and if users want to turn it off, they can do so by going to Gmail’s Settings > General and then turn off Grammar, Spelling, and Autocorrect.

Benefits

The benefits to users could be the ability to work smarter and faster, make a better impression with their communications, and learn more about correct spelling and grammar and use the knowledge to feel more confident in their writing over time.

Competition

These new features are, however, in competition with, and could affect profits for other third-party spelling and grammar services e.g. Grammarly which relies upon users opting to use an upgrade to a freemium service.

Texting To Blame?

Ofcom figures showing that teenagers, young adults and adults chose text-based communications as their preferred way of communication highlight one of the reasons why many people believe that mobile phone usage, with its reliance on auto-correct, is partly to blame for a decline in the standards of spelling and grammar that has necessitated a further need for more sophisticated AI-based tools.

What Does This Mean For Your Business?

For third-party spelling and grammar apps such as Grammarly, this move by Google could reduce the value of (and the perceived need for) their services.

For business users of Google’s services (G Suite and Gmail) these kinds of tools could help save time and improve the quality and consistency of their communications which in turn could positively reflect on their brands.

For Google, these new features could provide an improved experience for their users and add more loyalty and perceived value to Google’s range of services.

Some critics have, however, noted that these features could homogenise the way that we write, could raise privacy concerns, and that the AI technology could also be subject to bias in its suggestions.

Tech Tip – Split

Posted on by Andy Miller

If you’d like to be even more productive and be able to multi-task while using your iPhone or iPad, the ‘Split’ web browser app allows you to run two browser tabs side by side.

The app works in portrait and landscape views, and for each website that’s split in the browser you have a back button, you can bookmark a page, and you can open one of the websites into full-screen mode and simply tap to go back to the split-screen.

The Split app is free from Apple’s App Store, and similar split browser apps are also available for Android.

Your Password Can Be Guessed By An App Listening To Your Keystrokes

Posted on by Andy Miller

Researchers from SMU’s (Southern Methodist University) Darwin Deason Institute for Cyber-security have found that the sound waves produced when we type on a computer keyboard can be picked up by a smartphone and a skilled hacker could decipher which keys were struck.

Why?

The research was carried out to test whether the ‘always-on’ sensors in devices such as smartphones could be used to eavesdrop on people who use laptops in public places (if the phones were on the same table as the laptop) e.g. coffee shops and libraries, and whether there was a way to successfully decipher what was being typed from just the acoustic signals.

Where?

The experiment took place in a simulated noisy Conference Room at SMU where the researchers arranged several people, talking to each other and taking notes on a laptop. As many as eight mobile phones were placed on the same table as the laptops or computers, anywhere from three inches to several feet away. The study participants were not given scripts of what to say when talking, could use shorthand or full sentences when typing and could either correct typewritten errors or leave them.

What Happened?

Eric C. Larson, one of the two lead authors and an assistant professor in SMU Lyle School’s Department of Computer Science reported that the researchers were able to pick up what people were typing at an amazing 41 per cent word accuracy rate and that that this could probably be extended above 41 per cent if what researchers figured out what the top 10 words might be.

Sensors In Smart Phones

The researchers highlighted the fact that there are several sensors in smartphones that are used for orientation and although some require permission to be switched on, some are always on.  It was the sensors that were always switched on that the researchers were able to develop a specialised app for which could process the sensor output and, therefore, predict the key that was pressed by a typist.

What Does This Mean For Your Business?

Most of us may be aware of the dangers of using public Wi-Fi and how to take precautions such as using a VPN.  It is much less well-known, however, that smartphones have sensors that are always on and could potentially be used (with a special app) to eavesdrop.

Mobile device manufacturers may want to take note of this research and how their products may need to be modified to prevent this kind of hack.

Also, users of laptops may wish to consider the benefits of using a password manager for auto-filling instead of typing in passwords and potentially giving those passwords away.

Over A Million Fingerprints Exposed In Data Breach

Posted on by Andy Miller

It has been reported that more than one million fingerprints have been exposed online by biometric security firm Suprema which appears to have installed its standard Biostar 2 product on an open network.

Suprema and Biostar 2

Suprema is a South Korea-based biometric technology company and is one of the world’s top 50 security manufacturers.  Suprema offers products including biometric access control systems, time and attendance solutions, fingerprint live scanners, mobile authentication solutions and embedded fingerprint modules.

Biostar 2 is a web-based, open, and integrated security platform for access control and time and attendance, manage user permissions, integrate with 3rd party security apps, and record activity logs.  Biostar 2 is used by many thousands of companies and organisations worldwide, including the UK’s Metropolitan Police as a tool to control access to parts of secure facilities. Biostar 2 uses fingerprint scanning and recognition as part of this access control system.

What Happened?

Researchers working with cyber-security firm VPNMentor have reported that they were able to access data from Biostar 2 from 5 August until it was made private again on 13 August (Suprema were contacted by VPNMentor about the problem on 7th August).  It is not clear how long before 5 August the data had been exposed online.  The exposure of personal data to public access is believed to have been caused by the Biostar 2 product being placed on an open network.

In addition to more than one million fingerprint records being exposed, the VPNMentor researchers also claim to have found photographs of people, facial recognition data, names, addresses, unencrypted usernames and passwords, employment history details, mobile device and OS information, and even records of when employees had accessed secure areas.

VPNMentor claims that its team was able to access over 27.8 million records, a total of 23 gigabytes of data,

Affected

VPNMentor claims that many businesses worldwide were affected.  In the UK, for example, VPNMentor claims that Associated Polymer Resources (a plastics recycling company), Tile Mountain (a home decor and DIY supplier), and Medical supply store Farla Medical were among those affected.

It has been reported that the UK’s data protection watchdog, the Information Commissioner’s Office (ICO) has said that it was aware of reports about Biostar 2 and would be making enquiries.

What Does This Mean For Your Business?

For companies and organisations using Biostar 2, this is very worrying and is a reminder of how data breaches can occur through third-party routes.

In this case, fingerprint records were exposed, and the worry is that this kind of data can never be secured again once it has been stolen. Also, the large amount of other personal employee data that was taken could not only affect individual businesses but could also mean that employees and clients could be targeted for fraud and other crimes e.g. phishing campaigns and even blackmail and extortion.

The breach may have been avoided had Suprema secured its servers with better protection measures, not saved actual fingerprints but a version that couldn’t be reverse engineered instead, implemented better rules on databases, and not left a system that didn’t require authentication open to the internet.  Those companies that are still using and have concerns about Biostar2 may now wish to contact Suprema for assurances about security.

Facial Recognition at King’s Cross Prompts ICO Investigation

Posted on by Andy Miller

The UK’s data protection watchdog (the Information Commissioner’s Office  i.e. the ICO) has said that it will be investigating the use of facial recognition cameras at King’s Cross by Property Development Company Argent.

What Happened?

Following reports in the Financial Times newspaper, the ICO says that it is launching an investigation into the use of live facial recognition in the King’s Cross area of central London.  It appears that the Property Development Company, Argent, had been using the technology for an as-yet-undisclosed period, and using an as-yet-undisclosed number of cameras. A reported statement by Argent (in the Financial Times) says that Argent had been using the system to “ensure public safety”, and that facial recognition is one of several methods that the company employs to this aim.

ICO

The ICO has said that, as part of its enquiry, as well requiring detailed information from the relevant organisations (Argent in this case) about how the technology is used, it will also inspect the system and its operation on-site to assess whether or not it complies with data protection law.

The data protection watchdog has made it clear in a statement on its website that if organisations want to use facial recognition technology they must comply with the law and they do so in a fair, transparent and accountable way. The ICO will also require those companies to document how and why they believe their use of the technology is legal, proportionate and justified.

Privacy

The main concern for the ICO and for privacy groups such as Big Brother Watch is that people’s faces are being scanned to identify them as they lawfully go about their daily lives, and all without their knowledge or understanding. This could be considered a threat to their privacy.  Also, with GDPR in force, it is important to remember that if a person’s face (if filmed e.g. with CCTV) is part of their personal data, and the handling, sharing, and security of that data also becomes an issue.

Private Companies

An important area of concern to the ICO, in this case, is the fact that a private company is using facial recognition becasuse the use of this technology by private companies is difficult to monitor and control.

Problems With Police Use

Following criticism of the Police use of facial recognition technology in terms of privacy, accuracy, bias, and management of the image database, the House of Commons Science and Technology Committee has recently called for a temporary halt in the use of the facial recognition systems.  This follows an announcement in December 2018 by the ICO’s head, Elizabeth Dunham, that a formal investigation was being launched into how police forces use facial recognition technology (FRT) after high failure rates, misidentifications and worries about legality, bias, and privacy.

What Does This Mean For Your Business?

The use of facial recognition technology is being investigated by the ICO and a government committee has even called for a halt in its use over several concerns. The fact that a private company (Argent) was found, in this case, to be using the technology has therefore caused even more concern and has highlighted the possible need for more regulation and control in this area.

Companies and organisations that want to use facial recognition technology should, therefore, take note that the ICO will require them to document how and why they believe their use of the technology is legal, proportionate and justified, and make sure that they comply with the law in a fair, transparent and accountable way.

Robot Tuck Shops About To Hit U.S. College Campuses

Posted on by Andy Miller

San Francisco-based start-up, Starship Technologies, has announced that it will be putting food delivery robots that respond to phone app orders on 100 U.S. university campuses in the next 24 months.

The Bots

It has been reported that 25 to 50 of the (23Kg battery-powered, six-wheeled) Starship bots will be let loose on each campus, with the ability to roam around seven days a week, from 8 am to 2 am. The self-driving bots drive at 4 mph and use 10 cameras, radar, ultrasound sensors, GPS, computer vision and neural networks to process what they see in order to negotiate their way safely around a 4 km radius.

The bot’s cargo bay is mechanically locked during the journey and can only be opened by the customer with their smartphone app. The location of the robots is tracked, so that customer knows the exact location of their order and receives a notification at the time of arrival.

Food

The college campus robots will be delivering breakfast, snacks, and a variety of other food to students on campus.  Also, the app can take orders from local restaurants which the Starship bots will deliver to students on the campus for $1.99 per shipment, with Starship getting paid by the restaurant for making each delivery.

Benefits

The obvious benefits of the food delivery robots are that they can work whatever hours they are required all year round with no pay, no holiday and no need for breaks. Also, the Starship bots have an advantage over other local delivery services because the bots are small, manoeuvrable, know their way around the expansive campuses (thanks to pre-loaded, 3D maps), there are several bots working on one site, and they won’t need to be subject to any authorisation checks for being there.

Bigger Goals

Starship has bigger plans for the bots and is reported to have the goal of getting the bots onto college campuses across the US serving 1 million students.

Starship has also started a package delivery service in neighbourhoods and parts deliveries on business and industrial campuses using the bots.

What Does This Mean For Your Business?

Amazon has been making the news over the past couple of years with its delivery drones and ‘Scout’ delivery robots, and the well-funded start-up Starship ($40 million in new funding) has shown how it has been able to move quickly into a niche and join the growing delivery robot/drone industry.  For the robot and drone operating companies (Amazon, UPS, Google, Starship) these bots offer a way to reduce costs, avoid road congestion problems, avoid labour problems, and potentially deliver 24 hours a day all year round.  Users of bot and drone services can expect convenience, greater control over orders, and the novelty and fun of the delivery experience.

The benefits of drones and robots, however, may come at the expense of jobs, more of which are being taken away by the advance of technology-fuelled automation across many industries.

Apple Launches ‘Apple Card’

Posted on by Andy Miller

Apple has launched its ‘Apple Card’ in the US in partnership with Goldman Sachs and with processing by Mastercard.

Card

The Apple Card can now be applied for by customers in the US through the Wallet app on iPhone (iPhone 6 and later).  The physical laser-etched card, which is made of titanium and has a typically clean Apple design has no card number, no CVV security code, and no expiration date or signature on it.  Although you can buy using the card, the real Apple Card product is incorporated in the Wallet on the customer’s iPhone and works through Apple Pay. Apple says that the card can be used to make purchases in stores, in apps and on websites.

Advantages

Apple says that the Apple Card is built on simplicity, transparency and privacy and that it completely rethinks everything about the credit card. The main advantages of the Apple Card are:

  • There are no fees.
  • It gives instant cashback on purchases.  When you buy something on the Apple Card, you receive a percentage of your purchase back in Daily Cash every day, there’s no limit to how much you can get, and that cash goes right onto the Apple card it can be used it just like cash. Apple says that customers will get 2 per cent Daily Cash every time they use Apple Card with Apple Pay, and 3 per cent Daily Cash on all purchases made directly with Apple, including at Apple Stores, apple.com, the App Store, the iTunes Store and for Apple services.
  • It is secure.  There are no numbers on the card itself and using Apple Card through the iPhone means that it is covered by all the usual Apple Pay security features e.g. Face ID, Touch ID, unique transaction codes.
  • It offers much greater privacy.  Apple says that it doesn’t store the details of where you shop, what you bought, or how much you paid, and Goldman Sachs will not sell or share your spending data to any third-party. Also, Mastercard simply processes payments between parties on the global network.
  • The Apple Card shows you how to pay less interest.  For example, the Apple Card shows you a range of payment options and calculates the interest cost on different payment amounts in real-time.
  • The card can help you make more informed purchase choices.  For example, everything you buy gets a category (food, entertainment, shopping) and a colour-coded chart displays how much you’ve pending on each category.

Small Print Warning

This may all sound wonderful but some commentators have warned that when you sign up for the Apple Card you sign up to the standard agreement offered by Goldman Sachs.  Within this agreement is an arbitration clause that essentially means that you waive the right to make any claims, participate in a class action, or be heard in a court at trial for anything related to the agreement.

It is, however, possible to opt-out of the Goldman Sachs arbitration clause within 90 days after opening the account by contacting the company using messages, calling a toll-free number, or writing to a Philadelphia P.O. Box (Apple Card gives full instructions).

What Does This Mean For Your Business?

For other banks and credit companies that are still using traditional cards, this may represent a threat, as Apple, a trusted and globally known brand is offering something that appears to be more convenient, more secure, and has obvious instant cashback perks.

For Apple, this venture is a way that it can offer value and generate even deeper loyalty with and become more attached to the lives of its customers. This creates another important competitive advantage for the tech giant and allows it to gain a deeper understanding of its customers and their habits (even though it says it won’t share any information about those habits).

This also represents an opportunity for Apple to diversify at a time when its iPhone sales have been a bit flat and move towards the provision of services as well as hardware.

Tech Tip – Gallery Go

Posted on by Andy Miller

If you’ve been looking for a good gallery app for Android, Google has created an offline and compact, lite version of Google Photos that is uncluttered and easy to use.

The Gallery Go app works offline, so it doesn’t sync to a Google account (like Google Photos), but it only has two tabs at the bottom for pictures and folders, useful search tabs at the top, and very a user-friendly layout.

Gallery Go enables easy copying and moving photos between folders, you can create new folders, and it supports SD card.  The app also has automatic organisation so that each night, Gallery Go will automatically organise your photos to group by: People, Selfies, Nature, Animals, Documents, Videos and Movies.

Gallery Go is available from the Google Play Store.