AI Sector Daily Digest — June 23, 2026

1. Five Eyes agencies warn that frontier AI could compress cyber-risk timelines

• The U.S., UK, Canada, Australia and New Zealand said frontier AI models are expected to transform both offensive and defensive cyber capabilities, with the timeline measured in "months" rather than years.1
• The joint statement urged faster patching and the use of AI for defense, including earlier weakness detection and faster incident response.1
• Source: Reuters, June 22, 2026.

2. Baseten raises $1.5 billion at a $13 billion valuation

• Baseten, a California AI infrastructure startup co-founded by Australians, said it raised $1.5 billion in a round valuing the company at $13 billion.2
• The round was led by Sands Capital and Wellington Management; Blackbird VC said it made its largest-ever investment, without disclosing the amount.2
• Source: Reuters, June 23, 2026.

3. Reflection AI signs a major SpaceX compute deal

• Reflection AI said it signed a deal with SpaceX for extra capacity at the Colossus 2 data center; Reuters reported that CNBC put the payment schedule at $150 million per month from July 1, 2026, through 2029.3
• CNBC's reported terms would total about $6.3 billion if the deal runs to the end of the term, though either company can end it with 90 days' notice after the first three months.3
• Source: Reuters, June 22, 2026.

4. IBM brings OpenAI's cyber tools into enterprise security work

• IBM said it joined OpenAI's Daybreak Cyber Partner Program and will integrate OpenAI's frontier AI capabilities into enterprise security workflows.4
• The company also launched an application-security service built on Project Lightwell, the IBM and Red Hat initiative backed by a $5 billion commitment to secure open-source software.4
• Source: Reuters, June 22, 2026.

5. Indonesia drafts a national plan to put AI into priority government programs

• A draft presidential regulation seen by Reuters would direct ministries and regional governments to adopt AI from 2026 to 2029, including in Indonesia's $15 billion free-meals program.5
• The draft says AI could help design region-specific menus, monitor kitchen hygiene, predict food demand, detect irregularities and report risks such as biometric misuse, IP violations and deepfakes.5
• Source: Reuters, June 22, 2026.