Unintended consequences of artificial intelligence pose big risks: ISACA research
- Published: Tuesday, 10 December 2019 08:38
The evolving relationship between artificial intelligence, automation and humans is expected to create promising opportunities in the tech workforce while simultaneously presenting ‘sobering concerns’, according to new ISACA research.
ISACA’s ‘Next Decade of Tech: Envisioning the 2020s’ survey of more than 5,000 business technology professionals shows that respondents are significantly more optimistic about how technology advancements in the new decade will impact their career than they are about how it will impact society as a whole. 59 percent express optimism for the career ramifications compared to only 40 percent who are similarly upbeat about the overall societal impact.
From a workplace standpoint, respondents are optimistic that technological sea changes will both position their organizations for success and bolster their pay:
- Nearly 9 in 10 respondents (87 percent) say AI/machine learning will have a major or moderate impact on enterprises’ profitability;
- 58 percent expect that the evolving technology landscape will result in pay increases for tech professionals.
The nature of most technology roles is expected to be recalibrated by AI and increased integration of technology in the workplace, with 93 percent of respondents expecting an augmented workforce—or people, robots and AI working closely together—to reshape how some or most jobs are performed.
While AI/machine learning is identified as the most important enterprise technology of the next decade, followed by cloud platforms and big data, the potential downside of malicious AI attacks factors into the more pessimistic views for how society could be impacted by tech in the next decade. Only 50 percent of respondents think it is likely or very likely enterprises will give the ethical ramifications of AI deployments sufficient attention.
Whether through malicious or errant uses of AI, the potential consequences of misuse could be severe, with respondents indicating the highest levels of concern for AI attacks involving:
- Critical infrastructure (73 percent)
- Social engineering (58 percent)
- Autonomous weapons (56 percent)
- Attacks targeting the healthcare sector (56 percent)
- Data poisoning (55 percent).