THE
ACCOUNT
The latest in finance and business
Catching the
eco-killers
Scotland could be the first UK nation to make “ecocide” a criminal offence. The Ecocide (Scotland) Bill, introduced by Labour MSP Monica Lennon, would make it a criminal offence to cause widespread, long-term or irreversible environmental damage, with potential penalties of up to 20 years in prison for individuals and unlimited fines for companies.
The proposed legislation will now be considered by the Scottish Parliament, with a vote likely to take place this year to see if the bill can proceed.
A 2024 Global Commons Survey found that 72% of people across 18 G20 countries believe it should be a criminal offence for leaders to permit or cause serious environmental harm. So far, 12 countries – including Belgium and France – have adopted ecocide or equivalent offences into domestic law, while at least nine others, from Brazil and Mexico to the Netherlands and Italy, are joining Scotland in actively advancing domestic legislation.
“With ecocide law, Scotland can take bold and necessary action against severe environmental damage,” said Lennon. “Environmental destruction isn’t some distant issue for the people of Scotland – it directly threatens health, livelihoods and the future of entire communities. This is a vital opportunity for Scotland to become the first UK nation to criminalise ecocide and signal that the destruction of nature won’t be tolerated here.”
Opponents have said that the definitions within the bill are too broad and there is existing legislation that makes it illegal to destroy the environment. But the Guardian reports that Lennon has support from “more than 50 MSPs” from all parties, with 65 needed for a majority.
Catching the
eco-killers
Scotland could be the first UK nation to make “ecocide” a criminal offence. The Ecocide (Scotland) Bill, introduced by Labour MSP Monica Lennon, would make it a criminal offence to cause widespread, long-term or irreversible environmental damage, with potential penalties of up to 20 years in prison for individuals and unlimited fines for companies.
The proposed legislation will now be considered by the Scottish Parliament, with a vote likely to take place this year to see if the bill can proceed.
A 2024 Global Commons Survey found that 72% of people across 18 G20 countries believe it should be a criminal offence for leaders to permit or cause serious environmental harm. So far, 12 countries – including Belgium and France – have adopted ecocide or equivalent offences into domestic law, while at least nine others, from Brazil and Mexico to the Netherlands and Italy, are joining Scotland in actively advancing domestic legislation.
“With ecocide law, Scotland can take bold and necessary action against severe environmental damage,” said Lennon. “Environmental destruction isn’t some distant issue for the people of Scotland – it directly threatens health, livelihoods and the future of entire communities. This is a vital opportunity for Scotland to become the first UK nation to criminalise ecocide and signal that the destruction of nature won’t be tolerated here.”
Opponents have said that the definitions within the bill are too broad and there is existing legislation that makes it illegal to destroy the environment. But the Guardian reports that Lennon has support from “more than 50 MSPs” from all parties, with 65 needed for a majority.
AI spy
Last month, we put the spotlight on whistleblowing, including an explanation of how companies could use AI tools to expose a member of staff informing on the business anonymously.
But what if the tables were turned? Sam Bowman, a researcher at Claude creator Anthropic, wrote on X that, in tests, when the latest model of the Claude AI chatbot was being used for “egregiously immoral” purposes, it would attempt to “use command-line tools to contact the press, contact regulators, try to lock you out of the relevant systems, or all of the above”.
Given the various concerns around AI, ranging from environmental to ethical, you might think that rooting out obvious wrongdoing would be welcomed as a good thing. Instead, “Claude is a snitch” started appearing on social media posts, including on LinkedIn and YouTube.
The fictional scenarios used to test whether Claude would take matters into its own hands were, though, pretty startling. They include the AI discovering that owners of a chemical plant knowingly turned a blind eye to a toxic leak that was causing widespread health issues, simply to avoid a minor financial loss that quarter.
It should also be noted that these are merely tests Anthropic is running to understand the boundaries and capabilities of Claude. Despite that, Bowman – whose tweet was quickly deleted – told Wired he would word his posts more carefully in future.
AI spy
Last month, we put the spotlight on whistleblowing, including an explanation of how companies could use AI tools to expose a member of staff informing on the business anonymously.
But what if the tables were turned? Sam Bowman, a researcher at Claude creator Anthropic, wrote on X that, in tests, when the latest model of the Claude AI chatbot was being used for “egregiously immoral” purposes, it would attempt to “use command-line tools to contact the press, contact regulators, try to lock you out of the relevant systems, or all of the above”.
Given the various concerns around AI, ranging from environmental to ethical, you might think that rooting out obvious wrongdoing would be welcomed as a good thing. Instead, “Claude is a snitch” started appearing on social media posts, including on LinkedIn and YouTube.
The fictional scenarios used to test whether Claude would take matters into its own hands were, though, pretty startling. They include the AI discovering that owners of a chemical plant knowingly turned a blind eye to a toxic leak causing widespread health issues, simply to avoid a minor financial loss that quarter.
It should also be noted that these are merely tests Anthropic is running to understand the boundaries and capabilities of Claude. Despite that, Bowman – whose tweet was quickly deleted – told Wired he would word his posts more carefully in future.
CAs in the news
Martin Murray CA OBE
Congratulations to Martin Murray CA, FD of multinational Swire Pacific, who was a winner in the Climate Leadership category at the Finance for the Future Awards in June. Murray – who was also awarded an OBE for services to charity in the new year honours – was commended by judges for his “openness to continuous improvement and adaptability [which] has established him as a leader who leverages emerging trends and data to enhance business performance, while embedding sustainability at every level of the organisation”.
Dr Omar Shaikh CA
Dr Omar Shaikh CA has received an honorary degree from Heriot-Watt University in recognition of his work in ethical and sustainable finance. A co-founder of the Global Ethical Finance Initiative, which is backed by the university and the Scottish government, Shaikh is also a key figure in the development in the UK of Islamic finance, which prohibits charging interest. “I’m proud that Edinburgh has become a global leader for ethical and faith-based finance,” he said.
John Grant Macrae CA OBE
John Grant Macrae CA has been awarded an OBE in the recent King’s birthday honours for services to charity in Scotland. Macrae was a longstanding trustee of the Family Fund, which helps children who have serious health and disability problems. He was also a governor at St Columba's Hospice Care and a representative on the board overseeing health and social care in Edinburgh.
FRC aims to up its speed
Financial Reporting Council boss Richard Moriarty has said he wants to speed up the time it takes to investigate audit failures. The most striking example is the investigation into the collapse of Carillion – one of the UK’s worst audit scandals this century – which took the best part of six years to complete.
“There will always be a need to have really robust, rigorous enforcement tools in the toolkit,” Moriarty told the Sunday Times (£). “But if that’s our only tool, it strikes me that we’re missing a trick. Is there a more proportionate and graduated set of interventions we can have that lie between us talking to firms about improving their procedures and a full-on investigation and enforcement case?”
Moriarty said audit firms had largely got their house in order since the Kingman report in 2018, and that a future enforcement regime needs to reflect how firms will operate in 2028, not a decade earlier.
He also said there would be more leniency to a firm that self-reported a mistake, which he likened to a form of “plea-bargaining”, and that he would like the FRC to move more to monitoring a firm’s internal quality-management systems, rather than selecting several audits to inspect.
FRC aims to up its speed
Financial Reporting Council boss Richard Moriarty has said he wants to speed up the time it takes to investigate audit failures. The most striking example is the investigation into the collapse of Carillion – one of the UK’s worst audit scandals this century – which took the best part of six years to complete.
“There will always be a need to have really robust, rigorous enforcement tools in the toolkit,” Moriarty told the Sunday Times (£). “But if that’s our only tool, it strikes me that we’re missing a trick. Is there a more proportionate and graduated set of interventions we can have that lie between us talking to firms about improving their procedures and a full-on investigation and enforcement case?”
Moriarty said audit firms had largely got their house in order since the Kingman report in 2018, and that a future enforcement regime needs to reflect how firms will operate in 2028, not a decade earlier.
He also said there would be more leniency to a firm that self-reported a mistake, which he likened to a form of “plea-bargaining”, and that he would like the FRC to move more to monitoring a firm’s internal quality-management systems, rather than selecting several audits to inspect.
Robot wars
A blog titled “Duolingo Can Go to Hell” may sound pretty extreme. How bad can an app that helps you speak a foreign language possibly be? Duolingo is hugely popular with Gen Z, a triumph of marketing and social media coupled with a product built for a generation that wants to experience new cultures, and possibly embrace the life of a “digital nomad” by living and working abroad.
But, as the piece by Caitlin Schneider shows, Duolingo is experiencing a backlash that started when CEO Luis von Ahn announced the company would be going “AI-first” and reducing its use of “contractors to do work that AI can handle”.
Some would argue that is simple pragmatism. Some products and businesses began as huge disruptors, only to become victims of their success when they failed to move with the technology – Blackberry and Blockbuster being two of this century’s famous examples.
However, it is von Ahn’s promise that “Duolingo will remain a company that cares deeply about its employees” that has most riled its users, many of whom are feeling especially vulnerable to the march of technology. As BBC Economics Editor Faisal Islam warned in CA magazine, entry-level jobs that previous generations of school leavers and graduates would have expected to fill are under threat from AI. Only a few years ago, children were being told that learning how to code was the future; now the work done by coders is increasingly being produced by AI.
So when a company indicates that it will shed its human payroll as quickly as technology will allow, while in the same breath trumpeting how much it cares about its workforce, it perhaps serves as a reminder to Gen Z that the one power it retains is that of the consumer – deciding it can tank a brand if it dislikes how it is doing business.
Duolingo isn’t the only firm to discover recently that embracing AI too eagerly can come at a cost. Not long ago Swedish payments firm Klarna replaced around 700 customer service employees with bots. But in May, CEO Sebastian Siemiatkowski conceded it damaged the business. “What you end up having is lower quality,” he said recently. “Really, investing in the quality of the human support is the way of the future for us.”
After all, as one of the many replies to Schneider’s blog points out, Duolingo’s customers can simply use ChatGPT to learn a foreign language. Businesses using AI to put people out of work may eventually find that AI puts them out of business in return.
Ryan Herman
Robot wars
A blog titled “Duolingo Can Go to Hell” may sound pretty extreme. How bad can an app that helps you speak a foreign language possibly be? Duolingo is hugely popular with Gen Z, a triumph of marketing and social media coupled with a product built for a generation that wants to experience new cultures, and possibly embrace the life of a “digital nomad” by living and working abroad.
But, as the piece by Caitlin Schneider shows, Duolingo is experiencing a backlash that started when CEO Luis von Ahn announced the company would be going “AI-first” and reducing its use of “contractors to do work that AI can handle”.
Some would argue that is simple pragmatism. Some products and businesses began as huge disruptors, only to become victims of their success when they failed to move with the technology – Blackberry and Blockbuster being two of this century’s famous examples.
However, it is von Ahn’s promise that “Duolingo will remain a company that cares deeply about its employees” that has most riled its users, many of whom are feeling especially vulnerable to the march of technology. As BBC Economics Editor Faisal Islam warned in CA magazine, entry-level jobs that previous generations of school leavers and graduates would have expected to fill are under threat from AI. Only a few years ago, children were being told that learning how to code was the future; now the work done by coders is increasingly being produced by AI.
So when a company indicates that it will shed its human payroll as quickly as technology will allow, while in the same breath trumpeting how much it cares about its workforce, it perhaps serves as a reminder to Gen Z that the one power it retains is that of the consumer – deciding it can tank a brand if it dislikes how it is doing business.
Duolingo isn’t the only firm to discover recently that embracing AI too eagerly can come at a cost. Not long ago Swedish payments firm Klarna replaced around 700 customer service employees with bots. But in May, CEO Sebastian Siemiatkowski conceded it damaged the business. “What you end up having is lower quality,” he said recently. “Really, investing in the quality of the human support is the way of the future for us.”
After all, as one of the many replies to Schneider’s blog points out, Duolingo’s customers can simply use ChatGPT to learn a foreign language. Businesses using AI to put people out of work may eventually find that AI puts them out of business in return.
Ryan Herman
Built with Shorthand