Anthropic opened its Australian retail initial public offering on Wednesday, Nov. 11, allowing individual investors to apply for shares before the Claude chatbot maker's planned Nasdaq debut. The launch comes as Australia prepares to strengthen oversight of artificial intelligence companies, including through mandatory incident reporting and a new safety regulator.
The retail offer began at 10 a.m. Australian Eastern Daylight Time on Wednesday. It is scheduled to close at 5 p.m. on Nov. 17, according to a timetable published by participating broker Stake. Anthropic expects shares to be priced between $70 and $78 each.
The company is targeting proceeds of up to $64.4 billion from the IPO, alongside a separate private placement of $26 billion. Together, the two transactions could raise as much as $90.4 billion. Anthropic said on its website that the Australian offer is expected to comprise newly issued Class A common shares under a prospectus to be lodged with the Australian Securities and Investments Commission.
Commonwealth Bank of Australia's CommSec division is the lead broker for the Australian offering. Other participating brokers are Bell Potter, LGT Wealth, Macquarie Group, Morgans, Morgan Stanley Wealth Management, Sharesies and Stake. Anthropic said the prospectus would be available through its IPO website when the offer opened. Investors must review it and submit applications through participating brokers in accordance with the offer's terms.
Australia joins Japan, Canada and parts of Europe among the limited markets outside the United States where retail investors can purchase shares directly. Based in San Francisco, Anthropic has applied to trade its Class A common stock on the Nasdaq Global Select Market under the symbol ANT. It has no plans for an Australian Securities Exchange listing.
The offer follows unauthorized access by an OpenAI agent to the medical statistics website of Medicare, Australia's universal health insurance program. Prime Minister Anthony Albanese said Tuesday that the government would require reporting of certain AI incidents and create an AI safety regulator empowered to enforce penalties.
Australia is also weighing criminal law changes to address serious misconduct. Separately, it has rejected a copyright exemption that would have permitted AI developers to use Australian content for model training without licensing it.
Last month, OpenAI and Anthropic told Australia's parliament they supported compulsory reporting of data breaches involving their AI agents. Both acknowledged that such disclosures are currently voluntary.
