Archive for category: Blog

At least 16 companies have signed up to voluntary artificial intelligence safety standards introduced at the Bletchley Park summit.

They include companies from China and the UAE.

The signatories have committed to voluntarily “work toward” information sharing, “invest” in cybersecurity and “prioritise” research into societal risks.

However, the standards have faced criticism for lacking teeth due to their voluntary nature.

Speaking at a follow-up to the Bletchley Park event in Seoul, The UK’s technology secretary, Michelle Donovan, said the Seoul event “really does build on the work that we did at Bletchley”.

While it is good that there are now standards to be worked towards, Fran Bennett, the interim director of the Ada Lovelace Institute., has warned that while they remain voluntary  there is a risk that many companies will ignore them.

“It’s great to be thinking about safety and establishing norms, but now you need some teeth to it: you need regulation, and you need some institutions which are able to draw the line from the perspective of the people affected, not of the companies building the things, she said.”

There is no doubts that the launch of ChatGPT has stimulated what has been called an “arms Race” in AI.

“There are just not enough people who “understand how to make these systems, how to make them really perform, and how to solve some of the challenges going forward,” says Andrew Rogoyski, director of innovation at the Surrey Institute for People-Centred AI at the University of Surrey.

According to some one of the most tedious parts of the job is analysis of financial data, summarising documents, creating content, drafting documentation, advanced searching, analytics and insight and knowledge management.

“When you look at some of the tasks that auditors were doing – some of the boring, mundane tasks around churning data and manipulating it into a format where you can then actually do something valuable with it – that’s where artificial intelligence can play a really big part,” says Matthew Campbell, audit chief technology officer for KPMG UK.

According to a KPMG survey four in 10 senior audit professionals expect that the increased efficiency that AI can bring will lead to a reduction in the size of auditing teams.

This will mean that a junior accountant’s role is likely to change more towards systems management – overseeing AI powered software and databases – and relationship management, and away from the traditional reconciliations and ledgers work.

This will mean refocusing their training.

But the profession has had its problems, among them KPMG, fined £1.5m for failings in its 2019 audit of advertising agency M&C Saatchi.

The UK’s accountancy regulator, the Financial Reporting Council (FRC), carried out 19 investigations,  in the 2022-2023 financial year, slapping fines of £40.5 million on auditing firms and their clients for audit failings.

But greater use of AI may help with that.

AI is particularly good at spotting anomalies in vast amounts of data, making it useful when examining what may amount to billions of transactions by a client.

Despite the hype there is apparently very little evidence of improved productivity, according to the BBC.

“Between 1974 and 2008 the UK’s productivity – the amount of output you get per worker – grew at an annual rate of 2.3%. But between 2008 and 2020 the rate of productivity growth collapsed to around 0.5% per annum.”

One of the explanations is that people are getting side tracked into messaging friends on Whatsapp, watching videos on YouTube, arguing angrily on Twitter rather than concentrating on work.

But there are other explanations, one being that we are “not measuring the impact of technology properly. The second is that economic revolutions tend to be rather slow-burning affairs”.

One professor at the University of Cambridge argues that data isn’t collected in ways that would help with understanding and measuring productivity.

She gives the example of a business that has dispensed with in-house technical staff and moved to outsourcing and therefore on paper this business looks smaller.

Other academics argue that when we look back to previous industrial revolutions we telescope the time their innovations took to have an effect.

The conclusion is “The technology is seemingly not the problem, and in some respects it is not the solution either. High productivity growth will come only to those that learn how to use it best.”

You can read the full article here: https://www.bbc.co.uk/news/business-66233654

Earlier this year the government published new guidelines to help businesses to protect themselves from the risk of cybercrime, which can cost them millions.

It argued that cyber threats should be prioritised as a key business risk like financial and legal challenges and should be put on an equal footing with other threats like financial and legal pitfalls.

Businesses should ensure that they have detailed plans in place to respond to and recover from any potential cyber incidents.

They should also regularly test their plans’ resilience, include forma\l mechanisms for reporting incidents and should also ensure employees have adequate skills and awareness of cyber issues so they can work alongside new technologies in confidence.

Figures show almost one in three (32%) firms have suffered a cyber breach or attack in the past year,

Lindy Cameron, National Cyber Security Centre CEO, said: “Cyber security is no longer a niche subject or just the responsibility of the IT department, so it is vital that CEOs and directors understand the risks to their organisation and how to mitigate potential threats.”

Basic protections every business should have include:

• boundary firewalls and internet gateways — establish network perimeter defences, particularly web proxy, web filtering, co ntent checking, and firewall policies to detect and block executable downloads, block access to known malicious domains and prevent users’ computers from communicating directly with the Internet
• malware protection — establish and maintain malware defences to detect and respond to known attack code
• patch management — patch known vulnerabilities with the latest version of the software, to prevent attacks which exploit software bugs
• allow listing and execution control — prevent unknown software from being able to run or install itself, including AutoRun on USB and CD drives
• secure configuration — restrict the functionality of every device, operating system and application to the minimum needed for business to function
• password policy — ensure that an appropriate password policy is in place and followed
• user access control — include limiting normal users’ execution permissions and enforcing the principle of least privilege