IT body proposes that AI pros get leashed and licensed to uphold ethics
- Reference: 1708000151
- News link: https://www.theregister.co.uk/2024/02/15/bcs_ai_register_ethics/
- Source link:
So says industry body BCS – formerly the British Computer Society – which reckons licenses based on an independent framework of ethics would promote transparency among software engineers and their bosses.
"We have a register of doctors who can be struck off," said Rashik Parmar MBE, CEO at BCS. "AI professionals already have a big role in our life chances, so why shouldn't they be licensed and registered too?"
[1]
Chief execs and other managers in leadership roles, who are frequently not techies, are making key decisions about technology development, and these people must be "held accountable for using AI ethically," he added.
[2]
[3]
"If this isn't happening, the technologists need to have confidence in the whistleblowing channels available within organizations to call them out; for example, if they are asked to use AI in ways that discriminate against a minority group."
This could be [4]bias in hiring decisions , or related to [5]live facial recognition surveillance cameras that are being frequently deployed by police forces around the globe.
[6]
The importance of AI ethics was amplified by the [7]Post Office scandal , says the BCS boss, "where computer generated evidence was used by non-IT specialists to prosecute sub postmasters with tragic results."
For anyone not aware of the outrageous wrongdoing committed by the Post Office, it bought the bug-ridden Horizon accounting system in 1999 from ICL, a company that was subsequently bought by Fujitsu. Hundreds of local Post Office branch managers were subsequently wrongfully convicted of fraud when Horizon was to blame.
The Post Office systems scandal demands a critical response [8]READ MORE
The sorry saga, counted as among Britain's greatest miscarriages of justice, destroyed the lives of the local branch managers, with many left bankrupt and some driven to suicide.
[9]Damn, even the Pope thinks AI and autonomous weapons need reining in
[10]With all eyes on OpenAI, Meta drags its Responsible AI team to the recycle bin
[11]AI to replace 2.4 million jobs in the US by 2030, many fewer than other forms of automation
[12]Artificial General Intelligence remains a distant dream despite LLM boom
In a paper, " [13]Living with AI and emerging technologies: Meeting ethical challenges with professional standard ," the BCS's Ethics Specialist Group recommends:
Every technologist working in a high-stakes AI role should be a registered professional meeting independent standards of ethical practice, accountability, and competence.
Government, industry, and professional bodies should support and develop these standards together to build public confidence and create the expectation of good practice.
UK organizations should be required to publish their policies on ethical use of AI in any relevant systems – and those expectations should extend to leaders who are not technical specialists, including CEOs and governing boards.
The UK government should aim to support organizations to set world-leading ethical standards.
Georgina Halford-Hall, CEO at WhistleblowersUK and Chair of Strategy & Policy for the All Party Parliamentary Group (APPG) for Whistleblowing, said in a statement: "AI and the rapid advances being made in the technology sector have left whistleblowers open to abuse, bullying, harassment and victimization. We fully support BCS as it looks to build a transparent culture around ethical practice which gives tech professionals the confidence to challenge when those standards are not met." ®
Get our [14]Tech Resources
[1] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=2&c=2Zc5DQk-sXZ8HhC9tuKBv@QAAAYk&t=ct%3Dns%26unitnum%3D2%26raptor%3Dcondor%26pos%3Dtop%26test%3D0
[2] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=4&c=44Zc5DQk-sXZ8HhC9tuKBv@QAAAYk&t=ct%3Dns%26unitnum%3D4%26raptor%3Dfalcon%26pos%3Dmid%26test%3D0
[3] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=3&c=33Zc5DQk-sXZ8HhC9tuKBv@QAAAYk&t=ct%3Dns%26unitnum%3D3%26raptor%3Deagle%26pos%3Dmid%26test%3D0
[4] https://www.theregister.com/2024/01/23/nyc_ai_hiring_law_ineffective/
[5] https://www.theregister.com/2024/01/30/uk_law_makers_say_live/
[6] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=4&c=44Zc5DQk-sXZ8HhC9tuKBv@QAAAYk&t=ct%3Dns%26unitnum%3D4%26raptor%3Dfalcon%26pos%3Dmid%26test%3D0
[7] https://www.theregister.com/2024/02/01/fujitsu_finance_chief_apologises_horizon_inquiry/
[8] https://www.theregister.com/2024/01/22/opinion_column/
[9] https://www.theregister.com/2023/12/15/pope_francis_ai_treaty/
[10] https://www.theregister.com/2023/11/20/meta_responsible_ai/
[11] https://www.theregister.com/2023/09/06/generative_ai_jobs_forrester_report/
[12] https://www.theregister.com/2023/07/04/agi_llm_distant_dream/
[13] https://www.bcs.org/articles-opinion-and-research/living-with-ai-and-emerging-technologies-meeting-ethical-challenges-through-professional-standards/
[14] https://whitepapers.theregister.com/
BCS favours registration. Wow.
So,... the non-registered/non-licensed professionals are going to become the go-to crowd for the opposition? Since when does registration/licensing improve a moving target like ethics?
Anything is used for "good" and "bad". Which is which depends on your perspective.
Professional registration is an obvious thing to do - and would require individuals to sign up to ethical standards, which will evolve over time. Governments and large organisations should insist on using registered professionals to deploy AI systems as part of their duty to complying with whatever AI regulation eventually comes in.
It allows professionals to be struck off the register for ignoring their professional code of conduct, and is a sanction used in many other industries e.g. the BMA and the SRA.
It's not a solution on its own, but is a key element of a larger picture. Another key element is proper protection for Whistleblowers, which we do not have today. There is a Whistleblowing Bill going through Parliament which would create an Office of the Whistleblower with investigatory powers and a remit to work with regulators and professional bodies. The Bill includes criminal sanctions for retaliation against Whistleblowers - where appropriate.
Non sequitur
" The importance of AI ethics was amplified by the Post Office scandal, says the BCS boss, "where computer generated evidence was used by non-IT specialists to prosecute sub postmasters with tragic results." "
Since when did the horizon system make use of "AI"?
Quite apart from which, unless there are some pretty specific definitions of 'good' and 'bad' ethics in relation to 'AI', it will be impossible to apply objective judgement to the vast variety of applications and outcomes. Existing 'registrations' and 'codes of conduct' in the IT sphere have so far not demonstrably contributed to good practice, but have involved costs and hoop-jumping for practitioners to no great gain for anyone. Most practitioners behave ethically anyway, some don't, and those that don't rarely get held to account regardless of their 'registrations'.
Will Ethical Computing have problems with non-binary entities?
Would by any chance the CEO bonus be dependent on revenues?
Would the accreditation be based on any technical standards or simply ethical?
I nearly destroyed my keyboard as I was drinking coffee when I read the last bullet point.
"The UK government should aim to support organizations to set world-leading ethical standards."
Racket
client> Hi I need you to create me a model using this data and such and such criteria
loicensed UK AI expert> I am sorry as a loicensed AI expert I am not allowed to help with your request
client> Fair enough. I'll try someone else
offshore AI expert> What's your desired lead time and budget?
Load of
"where computer generated evidence was used by non-IT specialists to prosecute sub postmasters with tragic results."
You didn't need to be an IT expert to know that something fishy was going on.
"computer said X" serves as a great excuse to blame for one's incompetence and malice.
these people must be "held accountable for using AI ethically"
I'm not sure many people would like to be held accountable for what a LLM spits out, seeing as that will depend entirely upon what data it's trained on. Even if you review everything fed into it, and count the number of occurrences of each thing, you still won't be able to predict what it'll spit out, because it's just far too complex statistically. So there really can't be any such job as "ethical AI controller", you can only look at what it produces and decide to act on that or not. Seeing as you will never know how it decided to reach a particular conclusion, it could be argued that it would always be unethical to use it for any real life decision that affects somebody's life.
The thing is, BCS needs to sort itself out before trying to force people to join. I'm afraid to say that even politicians can spot a desperate and flailing society at 200 yards and you can't expect the politicians to save you and try and make you look relevant.
On a hobbyist basis, I'm interested in an obscure area of history and so I'm a member of the professional body for that area of history. Their journal has an academic rating as being International A1, which is the highest rank internationally for reputable peer reviewed journals. Their newsletter has it's own rating, and is significant enough to have to send a copy to the National Archives for permanent storage. It has in person and online meetings at nationally prestigious locations that are near utterly inaccessible to the public, with large collections of members who literally wrote the history books who are always happy to have a friendly chat (or meet up outside of the society for lunch to discuss things of mutual interest) and their meetings can have senior politicians, civil or crown servants, or the societies patrons from the royal family show up.
Now let's compare that to the BCS. The BCS doesn't have a journal, let alone a reputable peer reviewed one. Their newsletter is unrated because it's utterly irrelevant and if the National Archives archives a copy it's by accident. BCS is less relevant to the IT profession or society as a whole than The Daily Mail, and that's really an accomplishment, and not one I would consider a compliment.
The BCS charges over 3 times more than the rather more important other society i'm a member of, and yet if it delivers any benefit for the cost then it fails to mention it.
The suggestion that you join to obtain the post nominal initials MBCS (for Member, British Computing Society) is ridiculous and pompous, and doing so carries somewhat less prestige than advertising that you are a member of alcoholics anonymous. Every other society in the country expects Fellows to display post nominal initials, not simple members and the differentiation screams desperation for free marketing from its members rather than advertising competence. And if BCS exerts any real influence on the direction of IT legislation or anything else nationally then it's either not evident, or very evident that they need to be kept away from decision makers for the good of society.
So yes, you can put me down as "unimpressed" with BCS.
I read that as IT body proposes AI pros get lashes..
Call the moderatrix again.. This sounds like a good idea!