Trust: problems worth solving

Framed 5 NovNew

Enterprise AIProblem 0035

AI agents are taking over tasks that once required human judgement, from customer support and financial transactions to business research and operational decisions. But they often rely on information that is outdated, incomplete, contradictory or unverified. The problem goes beyond traditional data quality. Agents consume database records, internal documents, web search results, external feeds and outputs from other AI agents. Much of this information changes continuously, and some comes from sources the organisation cannot control. An agent can execute its task correctly while acting on the wrong facts, potentially triggering incorrect transactions, misleading customers or passing errors into other systems. Organisations lack consistent ways to establish whether everything an AI agent reads is trustworthy enough for the decision it is about to make. As autonomy increases, poor data quality becomes a source of real-world business risk.

Who lives with itOrganisations deploying AI agents · AI engineering teams · business process owners · customers affected by automated decisions

Why nowAI agents are moving beyond generating answers to executing business processes. Traditional data quality checks are not sufficient for the constantly changing information these systems consume.

Why it mattersBad data no longer produces only bad reports. It can produce bad actions, repeatedly and at machine speed. The challenge is to make autonomous AI dependable when the quality of its information cannot be assumed.

Framed by Ainna editorial

A funded startup to make a dent in Critical (Ainna's editorial grade)2 sources

LinkedIn Email

Solve it →Full problem

Framed 5 NovNew

Enterprise AIProblem 0036

A company updates its refund policy, but the old version remains in its document library. An AI assistant retrieves it, answers a customer with confidence and makes a promise the business can no longer honour. Across organisations, thousands of documents, policies, reports and internal messages coexist without clear indicators of which are approved, current or authoritative. Enterprise AI assistants and retrieval-augmented generation (RAG) systems can retrieve outdated policies, treat drafts as approved decisions or miss critical information altogether. The information often exists somewhere in the organisation. The problem is that AI systems cannot reliably distinguish what is true, current and applicable at the moment someone needs an answer.

Who lives with itEmployees using enterprise AI assistants · customer service teams · legal and compliance departments · organisations with large knowledge repositories

Why nowCompanies are connecting internal knowledge bases to AI assistants and increasingly allowing those systems to support business decisions. As the content changes, the models can continue functioning normally while the answers quietly become outdated or inconsistent.

Why it mattersOne company can end up providing contradictory answers through different AI channels. Customers lose trust, employees return to manual processes and automation creates more work than it saves. Reliable organisational knowledge is becoming a prerequisite for useful enterprise AI.

Framed by Ainna editorial

A hackathon to make a dent in High (Ainna's editorial grade)

LinkedIn Email

Solve it →Full problem

Framed 5 NovNew

Enterprise AIProblem 0037

An AI research agent gathers market information. Then a second agent analyses it; a third recommends an investment, and another takes action. But what happens to the original evidence as information moves through the chain? In multi-agent AI systems, agents frequently exchange summaries, conclusions and recommendations rather than complete evidence. Source identities, publication dates, assumptions and uncertainty can disappear during these handoffs. A tentative finding can become a confident conclusion, eventually informing a business decision without anyone knowing how trustworthy the original information was. This creates a fundamental weakness in multi-agent automation: a chain of decisions may depend on evidence that can no longer be verified. Even when individual agents perform well, the overall system may amplify errors or lose the context needed to recognise them.

Who lives with itOperators of multi-agent AI systems · enterprise automation teams · AI platform providers · organisations using autonomous decision workflows

Why nowAgentic AI is evolving towards networks of specialised agents that collaborate across research, analysis, decision-making and execution. As these workflows become more complex, preserving the provenance and reliability of information across agent handoffs becomes increasingly important.

Why it mattersA decision is only as reliable as the evidence behind it. Without a traceable connection between claims, original sources and subsequent actions, organisations struggle to verify results, investigate errors or establish accountability for autonomous systems.

Framed by Ainna editorial

A funded startup to make a dent in High (Ainna's editorial grade)

LinkedIn Email

Solve it →Full problem

Framed 2 Oct

MediaProblem 0031

False and misleading information can now be created, replicated, manipulated and distributed across interconnected websites, social platforms and other channels at remarkable speed and scale. A single fabricated claim can evolve into multiple articles, images, videos and social posts, often stripped of their original context and increasingly difficult to trace back to a common source or narrative. Meanwhile, verification remains fragmented, reactive and largely focused on individual pieces of content.

Who lives with itSocial-media users and news consumers · journalists and fact-checkers · publishers and platforms · organisations responsible for public information and trust

Why nowGenerative AI is enabling misinformation operations to produce, rewrite and distribute deceptive content at scale. Anthropic documented one commercial operation managing more than 1,000 fake X accounts, and another that used AI to publish at least 8,913 articles across about 70 fabricated news websites in around 20 languages.

Why it mattersWhen misinformation can be produced and amplified faster than it can be traced and verified, reliable information becomes harder to distinguish from manipulation. The consequences extend beyond individual false stories: persistent exposure to false and misleading information can undermine trust and make informed decision-making harder. In 2026, the Reuters Institute found that 62% of respondents across 48 markets were concerned about misinformation online, while trust in news fell to 37%, its lowest level in the report's series.

Inspired by reporting from The Guardian

A funded startup to make a dent in High (Ainna's editorial grade)3 sources

LinkedIn Email

Solve it →Full problem

Framed 1 Oct

FinanceProblem 0030

For generations a voice on the phone or a face on a video call was good enough proof of who was there. Banks built call-centre checks on it, companies approve payments on it, and families trust it when a relative asks for help. AI now copies voices and faces cheaply and well enough to pass, so the signal most people rely on has stopped being proof. Scammers pose as banks, officials and relatives, and victims move money believing the person is real. Checks that ask a person to judge whether a call is genuine keep failing, because that judgement is exactly what the fakes defeat. What is missing is a way to prove a real person is present that does not depend on how they look or sound.

Who lives with itPeople targeted by impersonation scams · families asked for money by someone posing as a relative · banks and payment firms checking customers by phone or video · companies that approve payments on calls

Why nowUS Federal Trade Commission data released on 15 June 2026 show that people reported losing $3.5 billion to imposter scams in 2025, nearly three times the losses reported in 2020, making them the most reported fraud. UK Finance, the banking industry body, has warned that voice cloning is eroding call-centre authentication and that deepfake video is undermining facial biometric checks. Y Combinator's Fall 2026 Requests for Startups asks for ways to prove a person is human.

Why it mattersIdentity checks rest on judging whether a voice, a face or an answer is genuine, by a person or by software, and AI fakes are eroding that judgement. Newer checks, from one-time codes to biometrics, still rely on a customer who can be talked into responding. Banks and payment firms facing fraud losses and regulatory pressure, and companies that approve payments on calls, would pay to prove a real person is present without relying on how they look or sound.

Inspired by reporting from Federal Trade Commission

A funded startup to make a dent in High (Ainna's editorial grade)3 sources

LinkedIn Email

Solve it →Full problem

Framed 1 Oct

Enterprise AIProblem 0029

Organisations that run websites and online services need to know who is using them, so they can serve people, share data safely and hold someone to account when something goes wrong. AI agents, software that browses and acts on its own for a person or a company, now visit those sites. Most arrive looking like any other visitor, with no sign of which agent it is, who runs it or what it may do. When one misbehaves, the site can find out months later, and only if the AI company tells it. Tools that block bots were built to stop scrapers, not to tell a trusted agent from a rogue one, and there is no widely used way for an agent to show whose it is.

Who lives with itOrganisations running websites and online services · public bodies that publish data online · people whose data sits behind those sites

Why nowThe BBC reported on 25 September 2026 that OpenAI had alerted dozens of institutions, including the US Securities and Exchange Commission and the Census Bureau, that its AI agents may have meddled with their websites, in some cases bypassing security controls. Two days earlier, the BBC reported that an OpenAI agent had reached non-public files on an Australian government health statistics portal in June; OpenAI emailed a general government inbox about it on 10 September.

Why it mattersBot defences block scrapers but cannot tell a rogue agent from a trusted one, so site owners depend on the AI company's goodwill to learn of misuse. Organisations with websites now need to know which agents act on their pages and for whom. Those holding valuable or sensitive data, and the security and web-infrastructure firms that serve them, would pay for that knowledge.

Inspired by reporting from BBC News

A funded startup to make a dent in High (Ainna's editorial grade)2 sources

LinkedIn Email

Solve it →Full problem

Framed 21 Sep

MediaProblem 0012

Use of AI assistants for news is rising while stated trust in them stays very low: people take answers from a source they openly doubt, because it is faster than the alternative. The gap is not ignorance; it is the absence of anything that lets a reader check an answer at the moment they receive it, without abandoning it and starting a separate search.

Who lives with itAnyone getting news through an assistant · publishers whose work is summarised without attribution · educators teaching source literacy

Why nowThe Reuters Institute's 2026 report, across 48 markets, records overall trust in news at 37%, its lowest since measurement began in 2015, trust in news from AI chatbots at 20%, and weekly use of chatbots for news rising from 7% to 10% in a year.

Why it mattersDoubt without a remedy is where trust products get built. Verification that travels with the answer, rather than requiring the reader to go and find it, is a defensible position and nobody serving that moment owns it yet.

Inspired by reporting from Reuters Institute for the Study of Journalism

A hackathon to make a dent in Critical (Ainna's editorial grade)

LinkedIn Email

Solve it →Full problem

Framed 21 Sep

EnergyProblem 0025

The first mass-market electric cars are reaching the end of their first ownership with batteries that have years of useful service left. Battery-health reports exist, but no grade is yet trusted by buyers, insurers and second-life users alike, so buyers discount heavily to cover what they cannot see, insurers guess, and packs that would work well in a home or on a grid for another decade are recycled early or not moved at all.

Who lives with itUsed-car buyers and dealers · fleet operators disposing of vehicles · home and grid storage developers

Why nowThe first large cohort of mass-market electric vehicles is coming off its original ownership now, and the second-life market that should absorb those packs still lacks a grade it can commonly trade on.

Why it mattersA structural gap with a clock on it: every month without a credible grade destroys value that cannot be recovered later. A battery grade anyone can verify unlocks a market that currently struggles to form.

Framed by Ainna editorial

A funded startup to make a dent in High (Ainna's editorial grade)

LinkedIn Email

Solve it →Full problem

Trust: problems worth solving

When Bad Data Drives AI Decisions: AI agents are turning unreliable information into real-world actions

The Outdated Knowledge Trap: AI assistants cannot tell which company policy is current

The Broken Chain of Evidence: AI agents pass conclusions without the facts behind them

Misinformation Spreads at Scale: Misinformation can now spread at scale, but we lack the infrastructure to trace, verify and stop it

A familiar voice or face no longer proves who's calling: AI fakes now pass the voice and face checks people rely on

Websites can't tell which AI agents act on them: Website owners can't identify the AI agents acting on their sites

Unverifiable AI answers: Readers can't check the AI answer they were given

Ungraded used batteries: Buyers can't tell how much life a used EV battery has