After I joined Forrester in 2019 and began protecting the bot administration software program market, bot administration software program helped firms reply a primary query (bot or not?) and guarded in opposition to bots. Later, these instruments helped reply a extra nuanced query (good bot, dangerous bot, or human?) and enabled people and good bots whereas defending in opposition to dangerous bots. At present, with the explosion of generative AI and AI brokers, companies want instruments that inform them whether or not utility site visitors comes straight from a human, from a bot, or from AI, in addition to what the intent of that site visitors is, to allow correct protections.
On the similar time, net utility firewall (WAF) distributors have consolidated their choices (which embody WAF, API safety, and bot administration) into net utility safety platforms, providing prospects a bundle of associated utility safety instruments with a unified knowledge mannequin and UI. As net utility safety platforms come collectively, the choice to go along with platform versus better of breed isn’t automated. We see some safety leaders depend on an internet utility safety platform for a lot of utility safety features after which add further layers of safety from distributors specializing in bots and AI brokers.
In brief, a brand new market has emerged. Bot administration use circumstances nonetheless matter and match effectively inside the purview of distributors which have added bot administration to their bundle of utility safety choices. However this new market, notably distributors which might be pushing the boundaries of buyer expertise and enterprise enablement use circumstances, should handle a wider vary of site visitors sources (suppose human prospects and their bots and AI brokers appearing on their behalf) and a broader set of use circumstances.
For instance, now this software program should tie AI agent site visitors again to a specific account and guarantee a constant and secure expertise for that account. It should present a seamless and trusted transaction expertise for patrons or something appearing on their behalf.
Finally, bot administration is simply too slender a reputation for the place this market goes. Beginning with our upcoming panorama report, we’re altering the title from bot administration software program to bot and agent belief administration software program. Forrester defines this as:
Software program that identifies and analyzes the intent of automated site visitors directed at an utility, establishing ongoing belief relationships with good bots and brokers and rejecting and misdirecting malicious bots and AI brokers to guard reliable buyer enterprise whereas additionally rising attacker prices.
Key factors about this title change:
It’s not nearly bots. Whereas AI agent site visitors continues to be a small fraction of web site visitors, we count on it to develop shortly. Because it does, organizations is not going to simply have to deal with bot site visitors however agent site visitors, too. The rise of AI brokers presents a problem to firms trying to perceive and handle buyer site visitors whereas nonetheless combating malicious utility layer assaults. It’s not sufficient to know whether or not inbound site visitors is from an AI agent; you could additionally know if the AI agent is appearing on behalf of a specific buyer or associate.
Belief strikes to the forefront. Bot administration has historically centered on defending organizations from a spread of enterprise logic assaults, comparable to account takeover, stock hoarding, advert fraud, and affect fraud. Whereas these situations nonetheless matter, they begin to fall below a broader banner of creating and constantly verifying trusted relationships with the vary of AI brokers and bots that interact together with your functions on behalf of consumers or as a part of formal and casual enterprise relationships. The choice isn’t “bot or not,” neither is it “good bot or dangerous bot” or “block or permit.” The choice is “How a lot do I belief this bot, AI agent, or human?” after which selecting actions primarily based on the diploma of belief.
I look ahead to sharing our panorama report on bot and agent belief administration software program on the finish of this yr. Within the meantime, please arrange an inquiry or steering session to debate how bot and agent belief administration software program can assist allow and shield your online business. I hope you too can be part of me at Safety & Danger Summit in November, the place you’ll hear extra about bots and brokers throughout a panel dialogue that I will probably be moderating with John Buten and Evelyn Mitchell-Wolf, “The Enterprise Of Bots, Scrapers, And AI Brokers.”
