KR — 전자신문
SAS: "Next Year Will Be the Year of AI Results… Accountability and Trust Within Companies Will Matter More"
AI2 min read
SAS, a data and AI software company, predicts that 2027 will be the year companies are judged on the real results of AI agents. It says trust, responsibility, and proof of value will matter most.
What happened
On the 17th, SAS released its '11 major AI industry predictions for 2027.' The company says AI agents, which are programs that can judge and act on their own, are spreading fast across business work. As a result, companies must rethink who is responsible, how they measure returns on AI investment, and how far they can trust AI decisions.
Key predictions
- SAS calls 2027 the year AI is judged by results, not promises.
- Chief information officers (CIOs) will manage the system access of AI agents, not just of people.
- In banking, AI agents may move from shopping to money transfers and other important decisions.
- Software makers must prove their products are worth buying, since AI makes it easier to build software in-house.
- Companies must find ways to measure whether AI output, such as content, code, and analysis, actually improves business results.
Why AI agents are different
Earlier AI tools mostly helped people with single tasks, like writing a draft. AI agents go further. They can plan steps and take actions in real systems. That is why questions about permission, safety, and who is to blame become urgent.
Why it matters
If an AI agent sends a wrong message to a customer or approves something without permission, the damage can reach the brand's reputation. SAS suggests companies must clearly define when humans step in and when an agent must be stopped. Simulations, digital twins (virtual copies of real systems), and synthetic data (artificial data used for testing) can help test risks before AI touches real customers. SAS Korea's CEO, Lee Jung-hyuk, said trust and accountability are most important in heavily regulated Korean industries such as finance, public services, manufacturing, and life sciences.
What to watch
- Whether companies set clear rules for AI agent permissions and responsibility.
- How banks check an agent's limits before it makes money transfers.
- Whether large cloud companies gain more power because AI infrastructure is so expensive.
- Whether companies start judging AI cost by total spending on data, governance, and infrastructure.
Glossary
- SAS
- A company that makes data analysis and AI software for businesses.