started · updated
Enterprise AI agents confront trust, governance and productivity gaps
A recent Forrester‑Boomi survey found that 86 % of enterprises have moved beyond testing AI agents and now use them in real business operations, yet only 34 % trust the decisions and actions of those agents. The study links the trust gap to insufficient data activation, governance and integration, noting that companies with strong AI controls avoid an average of $2.1 million in added costs from compliance penalties, customer loss, downtime and rework.
More than 60 % of organisations are already running AI agents in production workflows, and full autonomy without human review has risen sharply from 11 % to 26 %. At the same time, confidence in AI deployments among IT leaders fell from 40 % to 23 % over six months, reflecting growing concerns about governance and risk. Risk professionals see AI’s greatest value in verifying information, but 42 % report being directly affected by false or misleading AI‑generated content.
Gartner predicts that by 2028 AI agents will outnumber salespeople ten‑fold, yet fewer than 40 % of sellers say the agents improve productivity. Analysts warn that without a solid data foundation, workflow integration and seller‑experience design, organisations risk “agent sprawl” that adds digital activity without real performance gains. Effective orchestration, context management platforms and robust governance are cited as essential to turn AI agents into productive assets.
Entities
AI agents · AI agents · Argility Technology Group · Artificial intelligence · Billennium · Boomi · CJ Oosthuizen · Chief executive officers · Enterprise leaders · Gartner · Indermit Gill · International SOS
Claims
What the coverage asserts, and how many sources carry each claim.
- [○ 1 SOURCE] 86 % of enterprises have moved beyond testing AI agents and are using them in real business operations. backendnews.net
- [○ 1 SOURCE] Companies with strong AI governance avoid an average of $2.1 million in added costs from compliance penalties, customer loss, downtime and rework. backendnews.net
- [○ 1 SOURCE] Confidence in AI deployments among IT leaders fell from 40 % to 23 % over six months. www.itweb.co.za
- [○ 1 SOURCE] 42% of respondents say their organisation has been directly affected by false, misleading or unverified AI‑generated information. techafricanews.com
- [○ 1 SOURCE] Only 34 % of enterprises trust the decisions and actions of their AI agents. backendnews.net
- [○ 1 SOURCE] Nearly 60% of risk professionals view AI’s greatest value as verifying and validating information from multiple sources. techafricanews.com
- [○ 1 SOURCE] More than 60 % of organisations are running AI agents in production workflows. www.itweb.co.za
- [DISPUTED] 26% of enterprises give AI agents full autonomy without human review, up from 11% six months earlier. www.itweb.co.za
- [○ 1 SOURCE] Full AI autonomy without human review increased from 11% to 26% in a short period. www.itweb.co.za
- [○ 1 SOURCE] 42 % of organisations have been directly affected by false, misleading or unverified AI‑generated information. techafricanews.com
- [○ 1 SOURCE] By 2028 AI agents will outnumber salespeople ten‑fold, yet fewer than 40% of sellers say agents improve productivity. it-online.co.za
- [○ 1 SOURCE] 23 % of pharmaceutical firms have implemented agentic AI in 2024. zdrowie.co.pl