‘What is the Plan ready to invest in? Pornhub? ICE? Illegal detention centres?’
The Canada Pension Plan Investment Board's (CPPIB) approximately $416 million investment in xAI, Elon Musk's artificial intelligence company, is drawing criticism from governance experts and advocacy groups following a wave of controversies surrounding Grok, the controversial chatbot integrated into X, according to Canada’s National Observer.
The investment, disclosed in a quarterly report last year, funded xAI's development of large-scale, gas-powered data centres. However, the company has faced significant regulatory and reputational challenges since late December, when reports emerged that Grok was being used to generate non-consensual sexualized images, predominantly of women and children.
In a statement to Benefits and Pensions Monitor, CPPIB said they are closely following the recent controversy surrounding Grok’s image-generation tools and the regulatory attention it has triggered but still plan to maintain investment in xAI infrastructure.
"The misuse of Grok is deplorable and we denounce it. We unequivocally condemn the terrible practices that emerged in recent reports and take issues involving non-consensual imagery and child safety extremely seriously. These are matters for the operator’s governance, safeguards, and compliance. We treat the Grok incident pattern as a risk that must be corrected," said Michel Leduc, chief public affairs officer at CPPIB in an email.
"Our obligation is to deploy capital responsibly, manage risk rigorously, and hold counterparties to appropriate standards. Our financing is tied to the construction and operation of a data-center asset, part of xAI’s publicly reported Memphis expansion. We will continue to monitor developments and engage as appropriate within the scope of our role. For starters, we will push for hard blocks for non-consensual sexualized imagery and anything involving minors at generation time (not only after posting); independent safety assessment; transparency reporting; clear accountability between model team and platform moderation team, with board-level oversight; and incident response protocols," he added.
The fallout has since been swift as Indonesia and Malaysia banned the platform last week, while the UK's media regulator and California's attorney general have launched investigations. Canada's privacy commissioner has expanded an existing probe into xAI and X Corp to examine the platform's role in generating non-consensual deepfakes.
And for plan sponsors and fiduciaries monitoring CPPIB's investment decisions, the situation raises important questions about ESG integration and reputational risk management in alternative investments.
Eric Van Rythoven, a political science professor at Carleton University, characterizes the investment as "problematic," noting that content moderation concerns predated the current controversy. He told Canada’s National Observer that CPPIB may have underestimated the political risk exposure and could erode the fund’s value.
"This is a platform that is very volatile. There's a lot of exposure. There's a lot of political risk," Van Rythoven said. "One thing that CPPIB doesn't seem to be doing very well right now is accounting for the political backlash that Grok can face when it has no guardrails."
Charlie Angus, a former NDP MP and democracy advocate, questions whether CPPIB is applying appropriate investment standards given current geopolitical issues.
"It raises the question: what is the Canada Pension Plan ready to invest in? Pornhub? ICE? Illegal detention centres? Maybe the Canada Pension Plan could team up and invest in Russian bot farms," said Angus.
"There has to be a standard they're applying, particularly given the unprecedented threat that democracy and our sovereignty is facing. Otherwise they’re failing us,” he added.
Meanwhile, Patrick DeRochie, senior manager at Shift Action for Pension Wealth and Planet Health, suggests CPPIB may be positioning itself to profit from AI infrastructure development, including gas-powered facilities.
"It looks like the CPPIB is making a bet they can invest in the infrastructure behind AI: the data centres, the power generation, the pipelines that will supply that power," he said.
xAI announced last week that it has implemented “technological measures” to prevent certain image manipulations and will restrict features in jurisdictions where non-consensual deepfakes are illegal.
