Staff from the UK's leading AI labs raised concerns about the organization's governance and internal culture in a mouth-blowing complaint with charity Watchdog.
The Alan Turing Institute (ATI), a registered charity with considerable state funding, is under government pressure to overhaul its strategic focus and leadership after intervention from technology secretary Peter Kyle last month.
In a complaint to the Charity Committee, a current group of ATI staff members raised eight concerns and say the institute is at risk of collapse due to the government's threat to government funding.
The complaint alleges that the board, chaired by former Amazon UK boss Doug Gurr, failed to fulfill its core legal obligations, including providing strategic direction and ensuring accountability.
An ATI spokesman said the Charitable Committee had not contacted the Institute regarding complaints that could have been sent to the organization. They added that whistleblower complaints were filed last year with the government's UK research and innovation group funding ATI, and subsequent independent investigations determined there was no concern.
The complaint occurred after the ATI undergoing restructuring, with about 50 staff members (about 10% of the workforce) informing them that there was a risk of redundancy.
ATI's funding claims to cite “personally raised concerns” from an unknown industry partner, but warns that Kyle has revealed that future government support is conditional on improved provisions and change in leadership.
In this month's letter to Gurr, Kyle called for ATI's defense and national security-focused switch and leadership changes. The letter, stating that ATI “needs to receive the funds needed to implement reforms,” said it could review “long-term funding” next year.
The complaint alleges that there was no internal or external accountability for how the ATI fund was used. They claim that there is an internal culture of “fear, exclusion, defensiveness.”
The board also claims that it has not properly monitored the departure of a set of senior leaders under CEO Jean Innes, nor has it provided the appointment of senior leaders, claiming that it has “very undermined the credibility of staff, funders, partners and the wider public.”
The Guardian also learned that ATI has shut down online safety-related projects, tackling the housing crisis and reducing health inequality as part of a restructuring.
The restructuring has caused a dramatic change in the internal world at ATI, with over 90 staff members sending letters to the board last year warning that cost cuts are putting the organization's reputation at stake.
After the newsletter promotion
Some projects that are scheduled to close are working on developing AI systems to detect online harm, creating AI tools that will help policymakers tackle issues such as housing market inequality and affordability, and can measure the impact of health inequality in key policy decisions such as lockdowns.
Other projects that are expected to close include AI-based analysis of how government and the media interact. Projects considering social bias in AI outcomes will also be removed. The suspended projects include research into how AI affects human rights and democracy, and research into creating a global approach to AI ethics.
An ATI spokesman said: “We are forming a new phase in Turing, which requires major organizational changes to ensure the promise and unique role of the UK National Institute of Data Science and AI.
A spokesperson for the Charity Committee said the organization cannot confirm or deny whether it has received a complaint to protect the identity of whistleblowers.
