Most AI-powered cyber defence is only as effective as the data behind it, and a new report from Confluent finds that real-time data infrastructure remains a major gap for Singapore organisations trying to close it.
Key Takeaways
- Confluent’s 2026 Data Streaming Report finds 78% of Singapore IT leaders cite insufficient real-time data infrastructure as a barrier to AI adoption.
- 73% of Singapore IT leaders point to fragmented data ownership as a further obstacle to deploying AI-powered cyber defence.
- Confluent says real-time data streaming is becoming a core part of security architecture, as Singapore’s Cybersecurity Strategy 2026 calls for more proactive, shared defence against fast-moving threats.
The commentary, from Greg Taylor, Senior Vice President for APAC at Confluent, comes as Singapore’s recently announced Cybersecurity Strategy 2026 emphasises the need for more proactive defence as a shared responsibility against fast-moving cyber threats.
Isolated Signals Rarely Tell the Full Story
“Cybersecurity Awareness Month is a reminder that attackers are not waiting for organizations to catch up. AI is helping them move faster and scale techniques such as phishing and credential abuse, raising the pressure on defenders to respond at the same speed,” Taylor said. “AI-powered defense can help security teams detect anomalies, prioritize alerts and respond faster. But these tools are only as effective as the data behind them.”
Taylor noted that most cyberattacks do not appear as one obvious event: an unusual login, a suspicious IP address or an abnormal data transfer may mean little in isolation, but together can signal a developing threat. That makes timely, connected data critical for establishing what normal behaviour looks like and flagging deviations as they happen.
Infrastructure Gaps Are Slowing AI Adoption
Confluent’s 2026 Data Streaming Report found that 78% of Singapore IT leaders cite insufficient infrastructure to process real-time data as a challenge to AI adoption, while 73% point to fragmented data ownership. “When data is fragmented or delayed, it might be too late for teams spotting malicious activity only after damage has occurred,” Taylor said, adding that this is where real-time data streaming becomes part of the security architecture itself.
Humans Still Decide What Matters
Taylor was careful to frame the shift as augmentation rather than replacement. “While AI can surface patterns and reduce alert fatigue, it is still your workers who decide which risks matter, when to escalate and how to respond,” he said. “The future of cybersecurity is therefore not AI replacing human oversight. It is giving security teams trusted, real-time context so they can make faster, better-informed decisions before an anomaly becomes a breach.”
