UK Regulator Selects Major Banks to Test AI Applications Under New Oversight Program
The UK Financial Conduct Authority selected Barclays, Lloyds, and UBS for its AI Lab program, which allows firms to test real-world AI applications while regulators study the associated risks. The program is designed to help regulators understand AI before the technology outpaces existing rules. Florida also opened a criminal investigation into whether ChatGPT aided a gunman in a university shooting.
The UK Financial Conduct Authority selected Barclays, Lloyds, and UBS for its AI Lab program, which allows major banks to test real-world AI applications while regulators study the risks, according to reports from April 22, 2026.
The program is designed to help regulators understand how AI works in financial services before the technology outpaces existing rules. Banks will test applications including agentic AI systems and neurosymbolic models under regulated conditions.
The banking sector's cautious approach to AI oversight may set a template for regulators in other countries and industries.
In the United States, Florida Attorney General James Uthmeier opened a criminal investigation into whether ChatGPT provided advice that aided a gunman in a Florida State University shooting. Prosecutors are reviewing chat logs and have subpoenaed OpenAI for records related to its policies on threats, crime reporting, and training materials.
OpenAI said the chatbot only provided broadly available public information and did not encourage violence. The case pushes AI accountability into criminal law territory for the first time.
Anthropic's lobbying expenditures surged in the first quarter of 2026, surpassing OpenAI's spending. Policy debates around AI safety, compute access, cybersecurity, and national security are intensifying in Washington.
Insurers including QBE and Beazley are beginning to limit payouts for cyber losses and regulatory fines linked to AI use. Insurance markets are starting to treat AI as a distinct cyber risk category, which could influence how enterprises deploy AI systems.
Maryland Governor Wes Moore convened AI executives to discuss cybersecurity implications of advanced AI systems, reflecting growing concern among state leaders about AI's potential to outpace federal policy.
Meta is also collecting deeper workplace behavioral data from U.S.-based employees, including mouse movements, clicks, and keystrokes, to train AI agents capable of performing real-world tasks.
