Temasek International's investment chief Rohit Sipahimalani has identified the AI trade unwind and inflation as the two biggest risks facing global markets in 2027. Speaking at the Milken Asia Summit in Singapore, he noted that while an AI reversal may not be imminent, it could cause significant bumps along the way. Possible triggers include safety concerns, tighter regulation, or weak returns on AI investments.
The S$518 billion firm currently holds stakes in OpenAI, Anthropic, and Nvidia. Temasek plans to raise its AI allocation from 6% to up to 15% by 2031, but Sipahimalani wants to shift AI exposure from about 50% to 70 to 75 percent in public markets. This move would provide more flexibility to manage positions if market conditions change. He also flagged inflation and rising bond rates as a second key risk that could crack equity markets.
In corporate AI news, Penguin Solutions posted a strong fiscal Q4 2026 with net sales of $567 million, up 68% year over year. AI-driven businesses made up 78% of quarterly net sales, and Integrated Memory hit a record $341 million. Management projects fiscal 2027 net sales of about $2.43 billion. Meanwhile, Applied Digital reported first-quarter revenue of $341.9 million, up from $80.9 million a year earlier, as it expands data center and AI infrastructure investments.
Anthropic is asking select Claude users for permission to use their voice conversation data to train its AI models. The request covers audio recordings and associated conversation information from Claude's voice function, and users must give explicit consent as the authorization is deactivated by default. In healthcare, Oscar Health raised its 2026 outlook after strong Q2 results, reporting $4.9 billion in revenue and expecting material cost savings from expanded AI use in operations and claims processing.
California Governor Gavin Newsom signed three laws limiting AI authority in the workplace and schools. SB 947 requires managers to back up AI-driven firings with human evidence. AB 1979 prevents AI from replacing licensed professionals in healthcare. AB 1159 bars EdTech companies from training AI on student data, protecting about 2.9 million college students. Separately, Yoodli expanded its AI training tools for corporate use, reporting 900% revenue growth and about 100 employees.
Key Takeaways
- <ul><li>Temasek's investment chief names AI trade unwind and inflation as the top two risks for global markets in 2027.</li><li>Temasek, with S$518 billion in assets, holds stakes in OpenAI, Anthropic, and Nvidia.</li><li>Temasek plans to increase AI allocation from 6% to 15% by 2031 and shift public market AI share to 70-75%.</li><li>Penguin Solutions reported Q4 2026 revenue of $567 million, up 68% YoY, with AI driving 78% of sales.</li><li>Applied Digital revenue jumped to $341.9 million from $80.9 million a year earlier due to AI data center expansion.</li><li>Anthropic requests Claude user consent to use voice conversation data for AI model training.</li><li>Oscar Health raised 2026 guidance after Q2 revenue of $4.9 billion and expects AI cost savings in operations.</li><li>California signs three laws requiring human oversight for AI in firings, healthcare, and education.</li><li>SB 947 requires human evidence for AI-driven firings; AB 1159 protects 2.9 million students from AI training on their data.</li><li>Yoodli reports 900% revenue growth as it expands AI-powered corporate training tools.</li></ul>
Temasek warns AI unwind and inflation threaten 2027 markets
Temasek International's investment chief Rohit Sipahimalani named the AI trade unwind and inflation as the two biggest risks for global markets in 2027. He spoke at the Milken Asia Summit 2026 in Singapore on October 7. He said an AI reversal may not be imminent but could cause bumps next year. He also wants to move more of Temasek's AI holdings into public markets for flexibility.
Temasek CIO sees AI trade reversal as top market risk
Temasek's investment chief Rohit Sipahimalani called an AI trade unwind the biggest risk facing global markets. The S$518 billion firm holds stakes in OpenAI, Anthropic, and Nvidia. It plans to raise its AI allocation from 6% to up to 15% by 2031. He wants to shift AI exposure from about 50% to 70 to 75 percent in public markets for more flexibility. He also flagged inflation and rising bond rates as a second key risk for 2027.
Temasek investment chief flags AI and inflation for 2027
Rohit Sipahimalani, Temasek International's chief investment officer, said an AI trade reversal is the biggest risk for 2027. He spoke at the Milken Asia Summit 2026 in Singapore. He said possible triggers include safety concerns, tighter regulation, or weak returns. He also warned that inflation and higher bond rates could crack equity markets. He wants to lift Temasek's public market AI share to 70 to 75 percent. Temasek holds stakes in OpenAI, Anthropic, and Nvidia.
Penguin Solutions reports record Q4 sales driven by AI
Penguin Solutions posted fiscal Q4 2026 net sales of $567 million, up 68% year over year. Diluted EPS rose 133% to $1.00. AI-driven businesses made up 78% of quarterly net sales. Integrated memory hit a record $341 million in sales. Management projects fiscal 2027 net sales of about $2.43 billion and diluted EPS of $4.45.
Penguin Solutions stock jumps on strong AI earnings
Penguin Solutions stock jumped 13% to US$72.61 after strong Q4 2026 results. Revenue rose to about US$566.7 million from US$337.9 million a year earlier. Basic EPS climbed to US$1.46 from US$0.11. AI data center and memory products made up 78% of quarterly net sales. Integrated Memory brought in US$341 million, or 60% of company net sales, for the quarter.
Oscar Health raises 2026 outlook after strong Q2 earnings
Oscar Health reported strong Q2 2026 results with $1.1 billion in operating earnings, $1 billion in net income, and $4.9 billion in revenue. The company raised its full year 2026 guidance for revenue and operating earnings. Oscar Health is also expanding its use of AI in operations and claims processing, expecting material cost savings.
Anthropic asks Claude users for voice data to train AI
Anthropic is requesting permission from select Claude users to use their voice conversation data to train its AI models. The request covers both audio recordings and associated conversation information from Claude's voice function. Users must give explicit consent as the authorization is deactivated by default. The voice data permission operates independently from text conversation settings.
Editorial board calls for AI regulation on campus
An editorial board argues that artificial intelligence has no place on campus and should be regulated. The board urges administrations to examine AI use and consider potential risks and consequences. They believe society must be cautious about relying on AI and address concerns about its growing influence.
California laws require human oversight for AI decisions
California Governor Gavin Newsom signed three laws limiting AI authority in the workplace and schools. SB 947 requires managers to back up AI-driven firings with human evidence. AB 1979 prevents AI from replacing licensed professionals in healthcare. AB 1159 bars EdTech companies from training AI on student data, protecting about 2.9 million college students.
Yoodli expands into broader corporate training with new AI tools
Yoodli is repositioning itself as an AI platform for human skills with new tools that create courses, teach employees, and coach them during meetings. The company launched a new Create tool that builds training programs from prompts and expanded its Personas feature for answering questions. Yoodli has grown to about 100 employees and reported 900% revenue growth.
Applied Digital revenue jumps as AI investments expand
Applied Digital reported first-quarter revenue of $341.9 million, up from $80.9 million a year earlier. The company's net loss widened to $221.0 million due to higher spending on data centers and AI infrastructure. As of August 31, it held $3.7 billion in cash and $6.4 billion in debt. CEO Wes Cummins said the company is focused on building large-scale AI factory campuses with long-term contracts.
IDIRA webinar tackles AI-driven identity security risks
Palo Alto Networks is hosting a live webinar on securing enterprises from AI-driven identity risks. The session will cover reducing attack paths from AI and machine identities, controlling privileged access, and applying least privilege. It aims to help security teams manage risks as human, machine and AI identities converge. Registration is open through BankInfoSecurity.
icetana reports AI surveillance revenue growth and Antara sales
icetana Ltd saw recurring revenue rise 43% to $1.3 million in H1 FY26 and gross profit climb 56% to $1.15 million. The company made first sales of Antara Core and secured a $0.7 million deal with a new Australian customer. icetana's software uses machine learning to help security teams monitor large numbers of surveillance cameras. The company is expanding through partnerships in Japan, Asia-Pacific and the Middle East.
Sources
- Temasek chief investment officer says AI, inflation pose biggest market risks in 2027
- Temasek CIO Is Watching Two Risks to Global Markets in 2027, and AI Tops the List
- Temasek's investment chief warns AI trade reversal and inflation are top 2027 market risks
- Penguin Solutions (PENG) Fiscal Q4 2026 Earnings Call: AI Growth Drives Record Sales
- Penguin Solutions (PENG) Stock Jumps As AI Margins And Memory Scale Impress
- How Raised 2026 Outlook At Oscar Health (OSCR) Has Changed Its Investment Story
- Anthropic seeks Claude user voice data to train AI models
- Artificial Intelligence has no place on campus, its use should be regulated
- AI Can’t Fire You in California Without a Human
- Yoodli’s next act: AI that can teach workers almost any job skill
- Applied Digital's revenue soars, loss widens on growing AI investments
- Live Webinar | The Identity Imperative: Securing the Enterprise in the Age of AI-Driven Risk
- icetana Ltd (ASX:ICE): AI Surveillance Revenue Growth, Antara Sales and FY26 Commercial Progress
Comments
Please log in to post a comment.