Dive Brief:
- While generative AI entices enterprise leaders, the technology has challenges that block widespread use, according to a Forrester report published last week.
- Some generative AI tools have training data sets that are weighted toward publicly available internet data, which makes outputs reflect biases and misinformation. A lack of explainability compounds the problem, while privacy issues and regulatory concerns also impede enterprise adoption, according to Forrester.
- Internally, companies face more roadblocks to adoption. Nearly one-third of AI decision-makers say teams’ lack of technical skills is the biggest barrier to adoption, according to Forrester data. Other top concerns involve integrating generative AI with existing infrastructure and a lack of employee readiness.
Dive Insight:
Generative AI promises to accelerate enterprise operations, but companies must first pilot solutions, set metrics of success and adapt protocols to address potential risks.
Some businesses are standing by for more clarity on guidelines before jumping in, according to Forrester researchers.
“Many enterprises are waiting for more mature regulatory frameworks and clarity on the relevance of foundation models to specific industry verticals or corporate functions before putting genAI applications into production,” the firm said in the report.
But leaders are also feeling the pressure to adopt generative AI quickly to keep up with competitors. Early adopters are exploring internal and external use cases, each with its own list of pros and cons.
Software design, development and testing, for example, represent internal use cases for generative AI that can speed up code production. However, deploying these solutions could also result in faulty, plagiarized or insecure code, Forrester said in the report.
Despite the potential benefits of coding companions, more than half of organizations run across security issues with AI-generated code sometimes or frequently, according to a Snyk survey.
Forrester researchers recommend enterprises set standards for evaluating generative AI in vendor solutions, update AI strategies with guardrails and set governance guidelines for free versions of generative AI tools.
