BRICS Econ finds 68% prompt inconsistencies
- BRICS Econ published a report on August 6 saying enterprise vibe-coding certification now includes partners such as Scrimba and treats cross-platform prompt fluency as valuable. - BRICS Econ said 68% of users report prompt-engineering inconsistencies across platforms, a figure it used to argue for multi-tool enterprise training. (brics-econ.org) - Jansons Institute of Technology listed a “Prompt Engineering: From Beginner to Expert” webinar held on July 24, 2026. (jit.ac.in)
BRICS Econ said on August 6 that prompt engineering is being packaged less as a standalone specialty and more as part of broader enterprise “vibe-coding” certification. In a report on certification pathways, costs and governance, the group said enterprise programs now include partners such as Scrimba and argued that fluency across multiple prompting environments has become a practical requirement in company deployments. BRICS Econ said 68% of users report prompt-engineering inconsistencies across platforms. (brics-econ.org) The report used that figure to argue that enterprises need training that covers more than one model or toolset. (jit.ac.in) That framing matters because companies are moving from isolated chatbot use to workflows that span coding assistants, document tools and agent systems. BRICS Econ’s report presents prompt skills as part of operational governance — not only prompt writing, but also model selection, review controls and deployment standards. Scrimba, which BRICS Econ named as a partner example, separately lists prompt-engineering and AI-engineering course offerings on its site. ### Why is BRICS Econ focused on “inconsistencies across platforms”? The 68% figure is the report’s central data point. (brics-econ.org) BRICS Econ said users report different prompt behavior across platforms, which can affect repeatability when teams move the same task between tools or vendors. The report argues that this makes “multi-tool fluency” strategically valuable in enterprise settings, where outputs may feed into software, research or internal operations. OWASP’s GenAI security materials give one reason those differences matter in practice. (brics-econ.org) OWASP lists prompt injection as a leading risk category for LLM applications and says crafted inputs can alter a model’s behavior or output in unintended ways. That means enterprises are not only dealing with output quality differences, but also with security and control issues when prompts travel across systems. ### Where does Scrimba fit into this picture? Scrimba appears in BRICS Econ’s report as an example of the partner ecosystem around enterprise vibe-coding certification. Scrimba’s own site shows it offers prompt-engineering content and broader AI-engineering coursework, which aligns with BRICS Econ’s description of prompt skills being folded into wider technical training rather than sold as a single isolated credential. (brics-econ.org) Scrimba also said in a May 2026 article that standalone “prompt engineer” roles have declined while roles requiring prompt-engineering skills have increased. (genai.owasp.org) That is Scrimba’s characterization, but it points in the same direction as BRICS Econ’s report: employers may be embedding prompt work inside broader engineering and operations roles. ### Are schools and institutes still teaching prompt engineering directly? Jansons Institute of Technology said it held a webinar called “Prompt Engineering: From Beginner to Expert” on July 24, 2026. The event listing appears on the institute’s events page and shows that institutions are still running prompt-engineering sessions under that label even as industry reports describe the skill as part of larger AI-governance and deployment programs. (brics-econ.org) The institute’s listing does not provide a broader labor-market claim, but it does show that prompt engineering remains a named training category in academic and club programming. (scrimba.com) That leaves both tracks in view: direct instruction for learners and bundled certification for enterprise teams. ### What should readers watch next? OWASP published its current GenAI LLM Top 10 materials this week and continues to list prompt injection as a core risk area for LLM applications. BRICS Econ’s report suggests certification providers and enterprise buyers will keep tying prompt fluency to governance, security and cross-platform reliability. (jit.ac.in) Jansons Institute’s July 24 webinar page remains one public example of how that training is being presented to learners, while Scrimba’s course catalog shows where commercial providers are placing adjacent skills. (owasp.org)