Computer Weekly recently reported that in an interview with the outgoing England and Wales biometrics and surveillance camera commissioner, Professor Fraser Sampson, he warned of the declining state of oversight in AI facial recognition surveillance deployment by UK police.

Resignation 

Professor Fraser Sampson emailed his resignation letter to (then) Home Secretary Suella Braverman in August, stating his intention to resign by October 31. The reason given was that the Data Protection and Digital Information Bill will essentially make his role redundant by removing the responsibilities of the Biometrics Commissioner’s position and giving these powers to the Investigatory Powers Commissioner.

Professor Sampson, who was only appointed to the role in March 2021, said: “Having explored a number of alternatives with officials, I am unable to find a practical way in which I can continue to discharge the functions of these two roles beyond 1st November.” 

Professor Sampson’s responsibilities in the role had included overseeing how the police collect, retain and use biometric material (such as digital facial images), and encouraging their compliance with the surveillance camera code of practice.

Past Concerns and Criticisms 

In addition to espousing the many benefits of AI facial recognition’s deployment in the UK (e.g. catching known criminals – including those who are involved in child sexual abuse material), finding missing or vulnerable people, locating terror suspects, and helping to prevent the suffering of inhumane or degrading treatment of citizens, Professor Sampson has also previously criticised and raised concerns about aspects of its deployment. For example, in February, he noted:

– The absence of a clear set of legal rules or a framework to regulate the police’s use of AI and biometric material.

– A lack of clarity about the scale and extent of public space surveillance, particularly in relation to the proliferation of Chinese surveillance technology across the public sector.

Professor Sampson has also been vocal about a number of other related issues and concerns, such as:

– Issues related to the questionable legality of using public cloud infrastructure to store and process law enforcement data and the police’s general culture of retaining biometric data.

– Concerns about the unlawful retention of millions of custody images of people who have been charged with a crime. Despite Professor Sampson raising the issue, and the High Court ruling in 2012 that they should be deleted, it’s been reported that the Home Office, which owns UK police biometric databases, hasn’t done so because it has no bulk deletion capacity.

– The dangers of the UK slipping into becoming an “all-encompassing” surveillance state if concerns about these technologies (facial recognition) are not addressed. He has expressed his surprise at the disconnected approach of the UK government and his shock at how little the police and local authorities know about the capabilities and implications of the surveillance equipment they were using.

– Concerns about the possible misuse of facial recognition and AI technologies in controversial settings ( i.e. that the approach taken by UK police / their deployment methods in controversial settings could negate any benefits of the usage of the technologies). Controversial settings could include mass surveillance at public events, targeting specific communities, routine public surveillance, application in schools or educational institutions, and use in workplaces, all of which raise concerns about privacy, discrimination, and infringement on individuals’ rights.

– Rejection of the “nothing to worry about” defence, i.e. he challenged the common justification for surveillance that people who have done nothing wrong have nothing to worry about, stating this misses the point entirely.

– The government’s data reform proposals. For example, he criticised the government’s Data Protection and Digital Information (DPDI) Bill, arguing that it would lead to weaker oversight by subsuming biometric oversight under the Investigatory Powers Commissioner and removing the obligation to publish a Surveillance Camera Code of Practice.

– Efficacy and ethical concerns. Professor Sampson questioned the effectiveness of facial recognition in preventing serious crimes and highlighted the risk of pervasive facial-recognition surveillance. He also noted the chilling effect of such surveillance, where people might alter their behaviour due to the knowledge of being watched and warned against the abuse of these powers.

– He also advocated for a robust, clear, and intuitive oversight accountability framework for facial-recognition and biometric technologies, expressing concern about the fragmentation of the existing regulatory framework.

– The government’s lack of understanding and direction. For example, Professor Sampson commented on the lack of understanding and rationale in the government’s direction with police technology oversight and emphasised the need for public trust and confidence as a prerequisite, not just a desired outcome, for the rollout of new technologies.

– Predictive policing concerns. He warned against the dangers of using algorithms or AI for predictive policing, arguing that such approaches rely heavily on assumptions and create a baseline level of suspicion around the public.

Wider Concerns About Police Surveillance Using Facial Recognition 

Professor Sampson’s concerns about the police using Live Facial Recognition (LFR) surveillance at special assignments and high-profile events echo many of those expressed by others over the last few years. For example:

– Back in 2018, Elizabeth Denham, the then UK Information Commissioner launched a formal investigation into how police forces used FRT after high failure rates, misidentifications and worries about legality, bias, and privacy. In the same year, a letter written by privacy campaign group Big Brother Watch and signed by more than 18 politicians, 25 campaign groups, plus numerous academics and barristers, highlighted concerns that facial recognition is being adopted in the UK before it has been properly scrutinised.

– In the EU, in January 2020, the European Commission considered a ban on the use of facial recognition in public spaces for up to five years while new regulations for its use are put in place. In June this year, the EU actually adopted a blanket ban on AI-powered facial recognition in public spaces.

What Does This Mean For Your Business? 

The evolving landscape of the Data Protection and Digital Information Bill, particularly in the context of Professor Fraser Sampson’s resignation, could hold significant implications for UK businesses. This shift indicates a potential realignment of regulatory focus from physical biometric surveillance to digital data protection. For businesses, this underscores the need to adapt to a framework that prioritises digital data security and privacy.

The possible consolidation of regulatory bodies, like merging the roles of the Biometrics Commissioner into the Investigatory Powers Commissioner, may not necessarily suggest a decline of oversight, as suggested by Professor Fraser, but could also suggest a more streamlined oversight process. On the upside, this could mean simpler compliance procedures for businesses, but may also demand a broader understanding of a wider set of regulations. On the downside, companies (especially those dealing with biometric data) may need to very closely track these changes to ensure they remain compliant.

As the bill is likely to address the complexities of digital data, businesses will need to be proactive in understanding how these complexities are regulated. This is crucial for those handling large volumes of customer data or relying heavily on digital platforms. Adapting to evolving technologies and staying abreast of technological advancements will, therefore, be key.

All in all, in the light of the changes (and possible decline in oversight) highlighted by Professor Fraser, businesses will now need to be mindful of shifting political and public sentiments around privacy and surveillance, as these can influence consumer behaviour. While the changing regulatory landscape presents challenges, it also offers opportunities for businesses to align with contemporary data protection standards. Staying informed and adaptable may therefore be essential for navigating these changes successfully going forward.

Researchers at Bitdefender have reported discovering vulnerabilities in Google Workspace and Google Cloud Platform which, after first compromising the local machine, could allow threat actors to extend their activities to a “chain reaction” network-wide breach, potentially leading to ransomware attacks or data exfiltration.

The researchers say that, for example, starting from a single compromised machine, threat actors could: “move to other cloned machines with GCPW installed, gain access to the cloud platform with custom permissions, or decrypt locally stored passwords to continue their attack beyond the Google ecosystem.”

Bitdefender says it “responsibly disclosed” its findings to Google but says Google has confirmed “no plans” to address the findings, because it is outside of their specific threat model.

The advice to businesses is to strengthen detection and response capabilities (e.g. by investing in threat detection solutions, to identify and respond to unusual or unauthorised access attempts swiftly) and to have an incident response plan to address local device compromises.

Researchers at Bitdefender have reported discovering vulnerabilities in Google Workspace and Google Cloud Platform which, after first compromising the local machine, could allow threat actors to extend their activities to a “chain reaction” network-wide breach, potentially leading to ransomware attacks or data exfiltration.

The researchers say that, for example, starting from a single compromised machine, threat actors could: “move to other cloned machines with GCPW installed, gain access to the cloud platform with custom permissions, or decrypt locally stored passwords to continue their attack beyond the Google ecosystem.”

Bitdefender says it “responsibly disclosed” its findings to Google but says Google has confirmed “no plans” to address the findings, because it is outside of their specific threat model.

The advice to businesses is to strengthen detection and response capabilities (e.g. by investing in threat detection solutions, to identify and respond to unusual or unauthorised access attempts swiftly) and to have an incident response plan to address local device compromises.

Swedish startup Minesto has developed a subsea ‘kite’ style mini power plant that generates renewable energy from tidal streams and ocean currents.

How It Works 

The ‘wing’ technology, described by Minesto as a kind of “subsea kite” and a “powerful, lightweight, and modular power plant” which can be made with a wingspan ranging from 4.9 – 12m and weighing from 2.7 – 28 tonnes. Anchored to the seabed by a long cable tether, it sits in the sea and ‘flies’ across the main flow direction of the tidal streams and currents just like a kite flies in the air.

The wing technology uses the hydrodynamic lift force created by the underwater currents to move the kite around and its onboard control system autonomously steers the kite (using rudders and elevators) in a predetermined figure-of-eight trajectory. This has the effect of pulling kite and its turbine through the water at a flow that’s several times higher than the actual stream speed. This maximises the power it can generate and reduces the size of the kite and rotor required to collect the energy compared with a fixed turbine.

The turbine shaft inside the kite turns the generator which outputs the electricity to the grid via a power cable in the tether and a seabed umbilical to the shore.

Harnessing A Reliable And Inexhaustible Resource 

As highlighted by Minesto on its website, a balanced renewable energy mix is needed for the world to move towards a sustainable future energy system. Tidal streams and ocean currents are reliable and inexhaustible, available all over the globe, and are a rich source of energy that can be converted to a reliable and local source of renewable energy. This is why an easy to deploy and effective technology that can harness and use this endless resource (such as a simple kite system technology) could be a low cost and effective way to produce renewable (green) energy anywhere around the world (the ocean covers 71 per cent of the earth’s surface).

Predictable 

Also, unlike wind and solar, tidal streams and ocean currents are predictable, i.e. they’re caused by the gravitational forces exerted on the earth by the moon and are continuous and directional. This reduces risk and makes it easier in terms of control for the deployment of tidal power technology, such as Minesto’s wing/kite design.

Scalable 

The fact that the kites are a modular design which can be easily latched and unlatched (via the tether) to the seabed anywhere means that the system is easily scalable, simply by using hundreds of them across an area.

Real World Applications 

So far, Minesto reports that its wing subsea power generators have been delivering electricity to the Faroe Islands’ power grid since 2020 and, in 2022, Minesto commissioned the first power plant in Vetmannasund, Faroe Islands.

What Does This Mean For Your Organisation?

As Minesto rightly says, the world’s very necessary shift away from fossil fuels for power will involve developing and scaling a mix of innovative renewable energy solutions that make the most of existing natural resources such as wind, water, and sun. Also, with the UK being an island nation subject to tidal activity, in world where more roughly 71 per cent of the surface of the planet is covered by ocean, with its strong, constant, predictable tides, it does seem to be an area with the ability to supply vast amount of naturally generated energy if the right technology is deployed.

The advantages of the wing idea are that it can be easily and relatively cheaply deployed around the world, is scalable simply by multiplying the number used, can be placed far enough below the surface so as not to become a hazard or eyesore, and the technology is ready to go now. That said, these are relatively small turbines and even with many of them, there’ll still be a need for a mix of other ideas and solutions to harness the power of the waves.

These ideas will need to be part of wider mix of sustainable and renewable energy generating schemes that between them can offer enough power to seriously cut carbon emissions. Furthermore, they’ll need to supply the considerable energy needs of homes and businesses, provide power that’s affordable, have a low environmental impact, and thereby help the world to meet its climate targets as quickly as possible while still supporting the growth of the world’s economies.

As Minesto says, its wing solution ads a “step of energy conversion” that “expands the global tidal and ocean currents’ extractable potential.” 

If you need to search Google for a specific video, or for the best version of that video, Google Search lets you filter video results by duration, time, quality, source, and more. Here’s how it works:

– Type the words into Google Search that you’re looking for a video about.

– When the results are displayed, click on ‘Images’ (below the Google Search box).

– Click on the ‘Tools’ link (just to the right).

– Use the small dropdown menus to filter down and find the exact video you need.

One year on from its general introduction, OpenAI has announced some major upgrades to ChatGPT for its Plus and Enterprise subscribers.

New Updates Announced At DevDay 

At OpenAI’s first ‘DevDay’ developer conference on November 6, the company announced more major upgrades to its popular ChatGPT chatbot premium service. The upgrades come as competition between the AI giants in the new and rapidly evolving generative AI market is increasing, following a year that has seen the introduction of Bing Chat and Copilot (Microsoft), Google’s Bard and Duet AI, Claude (Anthropic AI), X’s Grok, and more. Although this year, ChatGPT has already been updated since its general basic release with a subscription service and its more powerful GPT-4 model, plug-ins to connect it with other web services, and integration with OpenAI’s Dall-E 3 image generator (for Plus and Enterprise) and image upload to help with queries, OpenAI will be hoping that the new upgrades will retain the loyalty of its considerable user base and retain its place as the generative AI front-runner.

GPT’s 

The first of four main new upgrades is ‘GPTs,’ which gives anyone (who is a ChatGPT Plus subscriber) the option to create their own tailored version of ChatGPT, e.g. to help them in their daily life, or to help with specific tasks at work, or at home. For example (as suggested by TechCrunch), a tech business could create and train its own GPT on its own proprietary codebases thereby enabling developers to check their style or generate code in line with best practices.

Users can create their own GPT with this ‘no coding required’ feature by clicking on the ‘Create a GPT’ option and using a GPT Builder. This involves using a conversation with the chatbot to give it instructions and extra knowledge, to pick what the GPT can do (e.g. searching the web, making images, or analysing data). OpenAI says the ability for customers to build their own custom GPT chatbot builds upon the ‘Custom Instructions’ it launched in July that let users set some preferences.

OpenAI has also addressed many privacy concerns about the feature by saying that any user chats with GPTs won’t be shared with builders and, if a GPT uses third party APIs, users can choose whether data can be sent to that API.

Share Your Custom GPTs Publicly Via ‘GTP Store’ 

The next new upgrade announced is the fact that users can publicly share the GPTs they create via a soon-to-be-launched (later this month), searchable ‘GPT Store’ – the equivalent of an app store, like Apple’s App Store or Google Play. OpenAI says the GPT Store will feature creations by verified builders and once in the store, GPTs become searchable and may “climb the leaderboards.” OpenAI also says it will spotlight the best GPTs in categories like productivity, education, and “just for fun,” and “in the coming months” GTP creators will be able to earn money based on how many people are using their GPT.

Turbo GPT-4 

In another announcement, OpenAI says it’s launching a preview of the next generation of its GTP-4 model (first launched in March) named GPT-4 Turbo.  As the name suggest, the Turbo version will be improved and more powerful. Features include:

– More up-to-date knowledge, i.e. knowledge of world events up to April 2023.

– A 128k context window to fit the equivalent of more than 300 pages of text in a single prompt.

– Optimised performance, which OpenAI says enables GPT-4 Turbo to be offered at a 3x cheaper price for input tokens and a 2x cheaper price for output tokens compared to GPT-4.

– ChatGPT Plus will also be easier to use, i.e. no need to switch between different models because DALL-E, browsing, and data analysis can all be accessed without switching.

Copyright Shield 

The last of the major update announcements for pro users is the introduction of ‘Copyright Shield’ to protect enterprise and API users (not free or Plus users) from legal claims around copyright infringement. This appears to be an answer to Microsoft’s September and Google’s October announcement that they will assume responsibility for potential legal risks to customers from copyright infringement claims arising from the use of their AI products.

Google, for example, announced it will offer limited indemnity and assume responsibility for the potential legal risks where customers receive copyright challenges through using generative AI products like Duet AI. Although it’s not yet clear how Copyright Shield will operate, OpenAI states in a recent blog: “we will now step in and defend our customers.” 

What Does This Mean For Your Business? 

OpenAI’s work with the other big tech companies and its general launch of ChatGPT a year ago have established it as the major player in the new and rapidly growing generative AI market. Building on the introduction of GPT-4 and rapid monetisation of its services through its business focused Plus and Enterprise subscriptions, these latest updates see OpenAI making the shift from AI model to developer to platform, i.e. with GTPs and the GTP Store.

What’s exciting and useful about GPTs is that they don’t require any coding skills, thereby democratising generative AI app creation and providing an easy way for businesses to create tools that can help them to save time and money, boost their productivity, improve their service, and much more. The addition of the GPT Store idea allows OpenAI to establish itself as a major go-to platform for AI apps, thereby competing with the likes of Google and Apple in a way. The Store could also provide a great opportunity for developers to monetise their GPTs as well as perhaps being a threat to consultancies and developers already creating custom AI services on behalf of paying clients.

The more powerful Turbo GTP-4 and its more up to date outputs, plus the lack of requirement to switch between different models are also likely to be features valued by businesses wanting easier, faster, and more productive ways to use ChatGPT. Furthermore, the Copyright Shield idea is likely to improve user confidence while enabling OpenAI to compete with Google and Microsoft, which have already announced their versions of it.

All in all, in the new and fast-moving generative AI market, these new upgrades see OpenAI ratcheting things up a notch, adding value, making serious competitive and customer retention efforts, showing its ambitions to move to platform status and greater monetisation, and further establishing itself as a major force in generative AI. For business users, these changes provide more opportunities to easily introduce customised and value-adding AI to any aspect of their business.