Menu
Subscribe to Holyrood updates

Newsletter sign-up

Subscribe

Follow us

Scotland’s fortnightly political & current affairs magazine

Subscribe

Subscribe to Holyrood
by Bill Magee
22 January 2024
Bug in the system: Questions remain over the use of AI in government

Artificial intelligence (AI) is playing a growing role in government | Alamy

Bug in the system: Questions remain over the use of AI in government

Surely the highlighting at Davos of the failings of artificial intelligent-generated systems will overcome any lingering reluctance by refuseniks to recognise what amounts to technological mismanagement on a global scale.

Of course, we're talking about AI's busy little cyber bugs - algorithms - causing digital chaos worldwide in both the public and private sectors. Especially when handling highly-sensitive data.

Add to this the protracted Post Office Horizon scandal with Fujitsu finally issuing an admission and apology - of sorts - over a faulty computer accounting system and leading to the ruination of the lives of several hundred postal workers and their families.

At the World Economic Forum in Switzerland OpenAI CEO Sam Altman, no less, warned his company expects to be "uncomfortable with some of the uses" of its tools, reports Axios, adding he admits such products will require "quite a lot of individual customisation".

Headline-grabbing Altman admits such algorithmic-laden future products could give different answers to different users based on their values and preferences possibly even what country they are in.

Highlighted is growing evidence of rising challenges associated with future-proofing and securing through trusted handling of a proliferation of data streams worldwide.

AI has increasingly been marketed as the panacea towards dealing with such an exponential rise in online and mobile data. Specifically algorithmic-led opaque lists of mathematical instructions and rules conducting specific tasks in hardware/software systems.

In the case of parliaments and the politicians within such hallowed institutions, in Scotland MSPs appear to have got the message in the wake of the SQA shambles when a parliamentary minute described the situation as the "tyranny of the algorithm".

Yet Whitehall appears unmoved. Specifically, the Cabinet Office where officials appearing algorithm-happy, as they controversially continue to delete massive amounts of files against advice from AI experts.

As recently as last October the UK Government was warned they continued to use poorly understood algorithms to make key decisions - leading to the practice being "uncontrolled and potentially discriminatory" in the way the technology is used.

It hasn't stopped Cabinet Office mandarins boldly proclaiming AI is "consistently more accurate than humans", according to media reports including publictechnology.net.

Excited civil servants are falling over themselves claiming to have "found an algorithm" they're busy employing to analyze, sort - also delete - a reputed 5.1 million of government documents, and counting.

For the tekkies among you: named an "Automated Digital Document Review" tool, it is being employed by the central government's Digital Knowledge and Information Management team. DKIM, of course, given Big Tech's obsession with acronyms. 

Apparently, the goal is to ascertain which ones need to be retained as the official record. Others are destroyed, deemed redundant, outdated, trivial, or ephemeral in nature or have reached their "retention period".

Surely such a mass document deletion jars with the very recent controversy involving certain top politicians and their WhatsApp messages disappearing.

New Scientist reports that on examination of current AI models, some already display individual "indicator properties" towards showing a conscience, "but there are no significant signs of consciousness."

However, it remains "sufficiently plausible" AI will become "more conscious" in the short term, warranting more investigation and preparation.

AI consciousness is not something any one discipline can tackle alone: "It requires expertise from the sciences of the mind, AI and philosophy."

If AI ever reaches such a state, the question is what sort of control will it have over humans and will it be able to manipulate us?

Worth bearing in mind when the Cabinet Office conducts its next annual review of the magic algorithm it has unleashed within the very nerve centre of the UK's political power base.

Holyrood Newsletters

Holyrood provides comprehensive coverage of Scottish politics, offering award-winning reporting and analysis: Subscribe

Read the most recent article written by Bill Magee - Cybersecurity: Why it’s becoming harder to stay safe online.

Get award-winning journalism delivered straight to your inbox

Get award-winning journalism delivered straight to your inbox

Subscribe

Popular reads
Back to top