AMD Forecasts AI Growth Alongside OpenAI and Meta Partnerships

AMD CEO Lisa Su is confidently betting on artificial intelligence, describing current investments as "the right gamble." She forecasts the AI data center market will swell to a staggering $1 trillion by 2030. AMD itself anticipates its revenue to climb 35% yearly, with its AI data center business projected to grow over 80% annually, reaching tens of billions of dollars by 2027. AMD is actively forging partnerships with industry giants like OpenAI, Oracle, and Meta, aiming to capture a substantial portion of this expanding market and compete effectively with companies such as Nvidia. Su underscores that increased computing power is a direct driver of innovation and productivity gains. However, the rapid proliferation of AI data centers brings significant environmental challenges. These massive facilities require immense amounts of energy and water for cooling. Researchers from Cornell University estimate that AI water consumption could hit 6 to 6.6 billion cubic meters by 2027, which is comparable to the annual water usage of 6 to 10 million American homes. Training AI models is particularly water-intensive; for instance, Mistral AI found that 91% of water use goes into training, with one page of text consuming as much water as growing a radish. This suggests that OpenAI's ChatGPT could be likened to a farmer harvesting a trillion radishes in terms of water footprint. Many data centers are being built in arid regions, straining local water supplies and depleting groundwater. Furthermore, these centers are expected to contribute 24 to 44 million tons of carbon dioxide to the atmosphere by 2030, equivalent to adding 5 to 10 million cars to the road. Experts suggest that "windbelt states" like Texas and Nebraska, or New York, might be more suitable locations due to their energy and water resources. Alongside environmental concerns, the widespread adoption of AI introduces new security complexities. Chief Information Security Officers (CISOs) are now tasked with AI security, often without the specialized knowledge required. A HackerOne report highlights that crowdsourced security, including bug bounties and red teaming, is proving highly effective, with 81% of CISOs finding it helpful for mitigating AI threats. This approach also helps address a global shortage of approximately four million cybersecurity professionals. Small and mid-sized businesses (SMBs) face heightened cyber risks due to inadequate AI governance, as cybercriminals increasingly leverage AI for sophisticated phishing and fraud schemes. Sean D. Mack from ISMG advises SMBs to establish clear AI policies, inquire about data handling practices, and bolster cyber hygiene and employee training. To counter these evolving threats, Legit Security has launched VibeGuard, a tool designed to secure AI-generated code directly within developers' integrated development environments. VibeGuard continuously monitors AI agents, preventing attacks and vulnerabilities from reaching production, and even trains AI agents to write more secure code, offering protection against risks like prompt injection attacks. Despite these challenges, AI is profoundly transforming various sectors. New AI-based "deep research" tools from companies like OpenAI and Google promise to dramatically accelerate research, with some, like Edison Scientific's Kosmos, claiming to condense six months of research into a single day. These tools can rapidly generate hypotheses and draft papers, though concerns about data fabrication and "hallucination" mean users must critically evaluate the reliability of AI-generated output. In the realm of drug discovery, BenchSci and Mila have partnered to leverage AI in biological research, aiming to expedite the development of new medicines and reduce associated costs and timelines. US intelligence analysts are also finding AI and automation indispensable for managing overwhelming data volumes, reducing analysis time from months to mere minutes, as demonstrated by AI-tagged images during the Ukraine conflict. To foster future AI literacy, Code.org and CSforALL are launching the "Hour of AI" during Computer Science Education Week, from December 8 to 14, 2025. This initiative offers over 100 free activities, contributed by more than 50 tech companies and nonprofits, to make artificial intelligence accessible and understandable for educators, families, and children through creative play and problem-solving.

Key Takeaways

• AMD CEO Lisa Su predicts the AI data center market will reach $1 trillion by 2030, with AMD's AI data center business growing over 80% annually to tens of billions by 2027.
• AMD is partnering with companies like OpenAI, Oracle, and Meta, aiming for a significant share of the AI market and competing with Nvidia.
• AI data centers are projected to consume 6 to 6.6 billion cubic meters of water by 2027 and add 24 to 44 million tons of carbon dioxide to the atmosphere by 2030.
• Training AI models, including OpenAI's ChatGPT, is highly water-intensive, with one page of text using as much water as growing a radish.
• CISOs are increasingly responsible for AI security, with 81% finding crowdsourced security effective for AI threats, helping address a global shortage of four million cybersecurity professionals.
• Small and mid-sized businesses face heightened cyber risks from AI due to weak governance, as cybercriminals leverage AI for phishing and fraud.
• Legit Security launched VibeGuard to secure AI-generated code, monitoring AI agents in development environments to prevent vulnerabilities and train for secure coding.
• AI-based "deep research" tools from companies like OpenAI and Google can significantly accelerate research but raise concerns about data fabrication and reliability.
• BenchSci and Mila are collaborating to use AI in biological research to speed up drug discovery and development.
• Code.org and CSforALL will host the "Hour of AI" from December 8-14, 2025, offering over 100 free activities to educate children and families about AI.

AMD CEO Lisa Su calls AI investment the right gamble

AMD CEO Lisa Su believes investing in AI is a smart move. She stated on CNBC's Squawk Box that the AI market will reach a trillion dollars by 2030. AMD expects its revenue to grow 35% yearly and its AI data center business to grow over 80% annually, reaching tens of billions by 2027. Su highlighted partnerships with companies like OpenAI, Oracle, and Meta. She emphasized that increased computing leads to more intelligence and productivity gains.

AMD CEO sees AI data market reaching 1 trillion by 2030

AMD CEO Lisa Su predicts the AI data center market will grow significantly. She believes it will reach 1 trillion dollars by the year 2030. Su discussed AMD's plan to lead in artificial intelligence growth on 'The Claman Countdown'.

AMD CEO Lisa Su defends AI investments as the right gamble

AMD CEO Lisa Su spoke about fears regarding AI investments, calling them "the right gamble". She believes more computing and investments lead to faster innovation. AMD updated its total accelerator market estimate to 1 trillion dollars by 2030, aiming for a large share. Su stated that the growing AI market favors innovation and partnerships, allowing AMD to compete with companies like NVIDIA.

AI data centers consume vast amounts of water

Artificial intelligence data centers use a lot of water for cooling servers. Cornell University researchers estimate AI water use could reach 6 to 6.6 billion cubic meters by 2027. Mistral AI found that training models uses 91% of their water, with one page of text using as much water as growing a radish. This means OpenAI's ChatGPT could be like a farmer harvesting a trillion radishes. Many data centers are built in dry areas, which strains local water supplies and can deplete groundwater.

AI data centers strain state power and water resources

AI data centers are huge buildings that need a lot of energy and water for cooling. Cornell University researchers found these centers will add 24 to 44 million tons of carbon dioxide to the air by 2030, like adding 5 to 10 million cars. They will also use as much water as 6 to 10 million American homes each year. States like Virginia and Texas are seeing development, but some centers are in water-scarce areas. Researchers suggest "windbelt states" like Texas and Nebraska, or New York, are better locations due to their energy and water resources.

CISOs now lead AI security with crowdsourced solutions

Chief Information Security Officers now have new responsibilities for AI security. Many CISOs lack the specialized knowledge needed for these threats. A HackerOne report shows that crowdsourced security is very effective, with 81% of CISOs finding it helpful for AI threats. This approach, including bug bounties and red teaming, helps find and fix vulnerabilities across software development. It also helps overcome the talent shortage in cybersecurity, where about four million more professionals are needed globally.

Poor AI rules increase cyber risks for small businesses

Small and mid-sized businesses face higher cyber risks due to weak AI governance. Sean D. Mack from ISMG warns that cybercriminals use AI for phishing and fraud. SMBs often adopt AI tools without proper policies or training. Mack advises businesses to create AI policies based on their risk and to ask about data handling before using any AI tool. He also suggests focusing on cyber hygiene and employee training to build stronger defenses.

AI research tools save time but raise trust concerns

New AI-based "deep research" tools from companies like OpenAI and Google can speed up research greatly. For example, Edison Scientific's Kosmos claims to do six months of research in one day. These tools can generate hypotheses and draft papers quickly. However, issues like data fabrication and "hallucination" mean users must question if they can truly trust the AI's output. The main concern is not if AI can research, but if it is reliable.

Hour of AI offers free activities to teach kids about AI

Code.org and CSforALL launched the Hour of AI with over 100 free activities. This event will happen during Computer Science Education Week from December 8 to 14, 2025. More than 50 tech companies and nonprofits contributed to the project. The goal is to make artificial intelligence easy to understand for educators, families, and children. It encourages learning through creative play and problem-solving.

Legit Security launches VibeGuard for AI code protection

Legit Security introduced VibeGuard, a new tool for securing AI-generated code. VibeGuard works directly within developers' AI-integrated development environments. It continuously watches AI agents, stopping attacks and preventing vulnerabilities before they reach production. The tool also trains AI agents to write more secure code. This helps security teams keep up with the fast pace of "vibe coding" and protects against risks like prompt injection attacks.

BenchSci and Mila partner to use AI for drug discovery

BenchSci and Mila announced a multi-year partnership to use AI in biological research. Their goal is to speed up drug discovery and development. Mila, a large deep learning research institute, and BenchSci, an AI reagent platform, will combine their knowledge. They plan to create new AI models to understand biological processes better and predict drug effects. This collaboration aims to lower the time and cost of bringing new medicines to patients.

AI helps intelligence analysts manage massive data

US intelligence analysts face a huge amount of data, making it hard to find important information. John Urby from GDIT says AI and automation are essential to handle this data overload. AI helps organize and label data, reducing analysis time from months to minutes. For example, AI tagged images in the Ukraine conflict, speeding up report declassification. While AI is a powerful partner, analysts still need to understand how AI reaches its conclusions and trust its output.

Sources

• AMD CEO Lisa Su: AI Investment is “The Right Gamble” in a Trillion-Dollar Market
• AMD CEO predicts AI data center market will grow to $1T by 2030
• AMD CEO Lisa Su Speaks Out Against Fears Surrounding AI Investments, Claiming It’s the “Right Gamble”
• Artificial Intelligence, Water Consumption and the Trillion-Radish Conundrum
• AI data centers could be coming to a city near you. Can your state handle it?
• Most CISOs now own AI security: Here’s what that means for your business
• Gaps in AI Governance Put SMBs at Greater Cyber Risk
• Mint explainer: Your deep research tool may save you time, but can it save you embarrassment?
• Hour of AI Unveils 100+ Free Activities to Help Demystify AI for Educators, Families, and Kids
• Legit Security Redefines Application Security for the AI Era With Launch of VibeGuard
• BenchSci and Mila Partner to Advance AI for Biological Inference and Accelerate Drug Discovery
• Too much data, too few analysts: How AI offers a ‘force multiplier’ for intelligence analysts